modelparams.dev

Every API parameter,
for every model.

An open, community-maintained catalog of LLM API parameters, scoped to the API surface your SDK calls. Choose a surface, then browse the exact request fields each model accepts.

Access type

Providers

Parameters

383 of 383 model surfaces

Bedrock

Claude Haiku 4.5 Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
Claude Opus 4.5 20251101 Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
Claude Opus 4.6 Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
Claude Sonnet 4.5 Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
Claude Sonnet 4.6 Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
DeepSeek R1 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
DeepSeek v3.2 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Devstral 2 123B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Gemma 3 12B IT Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Gemma 3 27B IT Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Gemma 3 4B IT Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GLM-4.7 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GLM-4.7 Flash Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GLM-5 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-OSS 120B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-OSS 20B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-OSS Safeguard 120B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-OSS Safeguard 20B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Kimi K2 Thinking Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Kimi K2.5 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Llama3 1 70B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Llama3 1 8B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Llama3 3 70B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Llama3 70B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Llama3 8B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Magistral Small 2509 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
MiniMax M2 Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
MiniMax M2.1 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
MiniMax M2.5 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Ministral 3 14B Instruct Converse API 7 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
additionalModelRequestFields.thinking.type
enum (disabled | enabled) "disabled" Controls Anthropic extended thinking on Bedrock. Bedrock rejects the adaptive mode the first-party Anthropic API accepts, so only disabled and enabled are listed here.
Budget tokens
additionalModelRequestFields.thinking.budget_tokens
integer (1024…+∞) Maximum token budget the model may use for extended thinking before producing the final answer. Must be less than inferenceConfig.maxTokens.
Only when additionalModelRequestFields.thinking.type = "enabled"
Ministral 3 3B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Ministral 3 8B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Mistral 7B Instruct Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Mistral Large 2402 Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Mistral Large 3 675B Instruct Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Mistral Small 2402 Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Mixtral 8x7b Instruct Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.top_k
integer (0…+∞) Limits token sampling to the top K most likely next tokens. Converse carries it through additionalModelRequestFields; it is not part of inferenceConfig.
Nemotron Nano 12B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Nemotron Nano 3 30B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Nemotron Nano 9B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Nemotron Super 3 120B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Nova 2 Lite Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.inferenceConfig.topK
integer (1…+∞) Limits token sampling to the top K most likely next tokens. Nova nests it under an inferenceConfig object inside additionalModelRequestFields.
Nova Lite Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.inferenceConfig.topK
integer (1…+∞) Limits token sampling to the top K most likely next tokens. Nova nests it under an inferenceConfig object inside additionalModelRequestFields.
Nova Micro Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.inferenceConfig.topK
integer (1…+∞) Limits token sampling to the top K most likely next tokens. Nova nests it under an inferenceConfig object inside additionalModelRequestFields.
Nova Pro Converse API 5 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Top K
additionalModelRequestFields.inferenceConfig.topK
integer (1…+∞) Limits token sampling to the top K most likely next tokens. Nova nests it under an inferenceConfig object inside additionalModelRequestFields.
Palmyra Vision 7B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Palmyra X4 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Palmyra X5 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Pixtral Large 2502 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Qwen3 32B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Qwen3 Coder 30B A3B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Qwen3 Coder Next Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Qwen3 Next 80B A3B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Qwen3 Vl 235B A22B Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Voxtral Mini 3B 2507 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Voxtral Small 24B 2507 Converse API 4 params
SDK method ConverseCommand API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
inferenceConfig.maxTokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
inferenceConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
inferenceConfig.temperature
number (0…1 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Bedrock rejects values above 1 for this model.
Top P
inferenceConfig.topP
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.

OpenAI

ChatGPT 4o Latest Chat Completions Retired 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-3.5 Turbo Chat Completions Deprecated 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-3.5 Turbo 16K Chat Completions 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4 Chat Completions Deprecated 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. OpenAI recommends sampling at 1.0.
Top P
top_p
number (0…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. OpenAI recommends sampling at 1.0.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
GPT-4 0613 Chat Completions Deprecated 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4 Turbo Chat Completions Deprecated 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4 Turbo 2024-04-09 Chat Completions Deprecated 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4.1 Chat Completions 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4.1 Mini Chat Completions 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4.1 Nano Chat Completions Deprecated 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4o Chat Completions 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4o 2024-11-20 Chat Completions 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-4o mini Chat Completions 3 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
GPT-5 Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5 Chat Latest Chat Completions Retired 1 param
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
GPT-5 Mini Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5 Nano Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.1 Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high) "none" Controls how much reasoning the model should perform before producing an answer.
GPT-5.1 Codex Max Responses API Subscription Retired 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.1 Codex Responses API Subscription Retired 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.2 Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.2 Codex Responses API Subscription Retired 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.2 Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.3 Codex Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.3 Codex Spark Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.3 Codex Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.4 Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.4 Mini Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.4 Mini Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.4 Nano Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.4 Pro Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.4 Pro Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.4 Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.5 Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.5 Pro Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
GPT-5.5 Pro Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.5 Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (minimal | low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed | none) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.6 Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "none" Controls how much reasoning the model should perform before producing an answer.
GPT-5.6 Luna Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "none" Controls how much reasoning the model should perform before producing an answer.
GPT-5.6 Luna Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (none | low | medium | high | xhigh | max) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.6 Sol Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "none" Controls how much reasoning the model should perform before producing an answer.
GPT-5.6 Sol Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (none | low | medium | high | xhigh | max) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-5.6 Terra Chat Completions 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high | xhigh) "none" Controls how much reasoning the model should perform before producing an answer.
GPT-5.6 Terra Responses API Subscription 3 params
SDK method responses.create Subscription
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (none | low | medium | high | xhigh | max) "medium" Controls how much reasoning the model should perform before producing an answer.
Reasoning summary
reasoning.summary
enum (auto | concise | detailed) "auto" Controls the level of reasoning summary returned with the response.
Output 1 param
Parameter Type Default Description Condition
Verbosity
text.verbosity
enum (low | medium | high) "medium" Controls how concise or detailed the model's final text response should be.
GPT-OSS 120B Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. OpenAI recommends sampling at 1.0.
Top P
top_p
number (0…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. OpenAI recommends sampling at 1.0.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
GPT-OSS 20B Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. OpenAI recommends sampling at 1.0.
Top P
top_p
number (0…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. OpenAI recommends sampling at 1.0.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
GPT-OSS Safeguard 120B Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. OpenAI recommends sampling at 1.0.
Top P
top_p
number (0…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. OpenAI recommends sampling at 1.0.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "medium" Controls how much reasoning the model performs when interpreting the provided safety policy before returning a classification.
GPT-OSS Safeguard 20B Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. OpenAI recommends sampling at 1.0.
Top P
top_p
number (0…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. OpenAI recommends sampling at 1.0.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "medium" Controls how much reasoning the model performs when interpreting the provided safety policy before returning a classification.
o1 Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
o1-mini Chat Completions Retired 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
o1-preview Chat Completions Retired 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high) "medium" Controls how much reasoning the model should perform before producing an answer.
o3 Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
o3-mini Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
o3-pro Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.
o4-mini Chat Completions Deprecated 2 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (16…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "medium" Controls how much reasoning the model should perform before producing an answer.

Anthropic

Claude 3.5 Haiku 20241022 Messages API Retired 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Claude 3.5 Haiku Latest Messages API Retired 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Claude 3.5 Sonnet 20241022 Messages API Retired 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Claude 3.5 Sonnet Latest Messages API Retired 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Claude 3.7 Sonnet 20250219 Messages API Retired 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude 3.7 Sonnet Latest Messages API Retired 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude 3 Opus 20240229 Messages API Retired 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Claude 3 Opus Latest Messages API 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Claude Fable 5 Messages API 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (adaptive) Only adaptive thinking is supported; omit the parameter entirely to run without thinking (an explicit disabled value is rejected).
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Fable 5 Messages API Subscription 4 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (adaptive) Only adaptive thinking is supported; omit the parameter entirely to run without thinking (an explicit disabled value is rejected).
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Haiku 4 Messages API 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Haiku 4.5 Messages API 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Haiku 4.5 20251001 Messages API 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Haiku 4.5 20251001 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Haiku 4.5 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Haiku 4 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Opus 4.1 20250805 Messages API Deprecated 7 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Claude Opus 4.1 20250805 Messages API Subscription Deprecated 7 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Claude Opus 4 20250514 Messages API Retired 7 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled"
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Claude Opus 4 20250514 Messages API Subscription Retired 7 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled"
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Claude Opus 4.5 20251101 Messages API 8 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 4 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Effort
output_config.effort
enum (low | medium | high) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.5 20251101 Messages API Subscription 8 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 4 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Effort
output_config.effort
enum (low | medium | high) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.6 Messages API 8 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"enabled", "adaptive"} or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"enabled", "adaptive"} or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"enabled", "adaptive"}
Reasoning 4 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type ∈ {"adaptive", "enabled"}
Effort
output_config.effort
enum (low | medium | high | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.6 Messages API Subscription 8 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"enabled", "adaptive"} or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"enabled", "adaptive"} or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"enabled", "adaptive"}
Reasoning 4 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type ∈ {"adaptive", "enabled"}
Effort
output_config.effort
enum (low | medium | high | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.7 Messages API 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "disabled" Controls the Anthropic thinking mode values supported by this model.
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.7 Messages API Subscription 4 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "disabled" Controls the Anthropic thinking mode values supported by this model.
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.8 Messages API 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "disabled" Controls the Anthropic thinking mode values supported by this model.
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4.8 Messages API Subscription 4 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "disabled" Controls the Anthropic thinking mode values supported by this model.
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 4 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Opus 5 Messages API 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "adaptive" Controls the Anthropic thinking mode values supported by this model.
Not when output_config.effort ∈ {"xhigh", "max"}
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Opus 5 Messages API Subscription 4 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "adaptive" Controls the Anthropic thinking mode values supported by this model.
Not when output_config.effort ∈ {"xhigh", "max"}
Thinking display
thinking.display
enum (summarized | omitted) "omitted" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Sonnet 4 20250514 Messages API Retired 7 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled"
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Claude Sonnet 4 20250514 Messages API Subscription Retired 7 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled"
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "enabled"
Claude Sonnet 4.5 Messages API 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Sonnet 4.5 20250929 Messages API 6 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Sonnet 4.5 20250929 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type = "enabled" or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type = "enabled" or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Sonnet 4.5 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Sonnet 4.6 Messages API 8 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"enabled", "adaptive"} or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"enabled", "adaptive"} or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"enabled", "adaptive"}
Reasoning 4 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type ∈ {"adaptive", "enabled"}
Effort
output_config.effort
enum (low | medium | high | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Sonnet 4.6 Messages API Subscription 8 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"enabled", "adaptive"} or top_p ≠ null
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"enabled", "adaptive"} or temperature ≠ null
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"enabled", "adaptive"}
Reasoning 4 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type ∈ {"adaptive", "enabled"}
Effort
output_config.effort
enum (low | medium | high | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Sonnet 4 Messages API Subscription 6 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when thinking.type ∈ {"adaptive", "enabled"}
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens whose cumulative probability reaches this value.
Not when thinking.type ∈ {"adaptive", "enabled"} or temperature ≠ 1
Top K
top_k
integer (0…+∞) 0 Limits token sampling to the top K most likely next tokens.
Not when thinking.type ∈ {"adaptive", "enabled"}
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive | enabled) "disabled" Controls the Anthropic thinking mode values supported by this model.
Budget tokens
thinking.budget_tokens
integer (1024…+∞) 4096 Maximum token budget Anthropic may use for extended thinking before producing the final answer.
Only when thinking.type = "enabled"
Claude Sonnet 5 Messages API 4 params
SDK method messages.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "disabled" Controls the Anthropic thinking mode values supported by this model.
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.
Claude Sonnet 5 Messages API Subscription 4 params
SDK method messages.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | adaptive) "disabled" Controls the Anthropic thinking mode values supported by this model.
Thinking display
thinking.display
enum (summarized | omitted) "summarized" Controls whether Anthropic returns summarized or omitted thinking content.
Only when thinking.type = "adaptive"
Effort
output_config.effort
enum (low | medium | high | xhigh | max) "high" Controls Anthropic response thoroughness and token spend.

Alibaba

DeepSeek v3.2 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
DeepSeek v4 Flash Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
DeepSeek v4 Flash 0731 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
DeepSeek v4 Pro Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
DeepSeek v4 Pro 0813 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
GLM-5.1 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
GLM-5.2 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Kimi K2.7 Code Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen Flash Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen Max Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen Plus Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen Turbo Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 235B A22B Thinking 2507 Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 30B A3B Instruct 2507 Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 30B A3B Thinking 2507 Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 Coder Flash Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Qwen3 Coder Next Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Qwen3 Coder Plus Chat Completions Deprecated 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Qwen3 Max Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 Next 80B A3B Instruct Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 Next 80B A3B Thinking Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 Vl 235B A22B Instruct Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3 Vl 235B A22B Thinking Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean false Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.5 Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.5 122B A10B Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.5 27B Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.5 35B A3B Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.5 397B A17B Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.5 Flash Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.
Reasoning 1 param
Parameter Type Default Description Condition
Enable thinking
extra_body.chat_template_kwargs.enable_thinking
boolean true Controls Qwen3 thinking mode when using OpenAI-compatible clients that pass provider-specific extra body fields.
Qwen3.6 27B Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.6 35B A3B Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.6 Flash Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.6 Max Preview Chat Completions Deprecated 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.6 Plus Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.7 Flash Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.7 Max Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.7 Plus Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.8 2.4t A95B Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.8 27B Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
Qwen3.8 Max Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, including both reasoning and the final answer.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (0…+∞) 20 Limits generation to the selected number of highest-probability tokens. Values above 100 disable top-k sampling.
Reasoning 2 params
Parameter Type Default Description Condition
Enable thinking
extra_body.enable_thinking
boolean true Toggles the model's hybrid thinking mode, sent as a provider-specific extra body field on OpenAI-compatible clients.
Thinking budget
extra_body.thinking_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before it starts the final answer; defaults to the model's maximum reasoning length.
Only when extra_body.enable_thinking = true
QwQ Plus Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.9 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
extra_body.top_k
integer (1…+∞) 20 Limits generation to the selected number of highest-probability tokens.

Google

Gemini 2.5 Flash Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (-1…24576) -1 Number of thinking tokens Gemini should use; 0 disables thinking and -1 uses dynamic thinking.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Flash Lite Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer 0 Number of thinking tokens Gemini should use; -1 uses dynamic thinking, 0 disables thinking, and fixed budgets start at 512 tokens.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Flash Lite Generate content Subscription 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer 0 Number of thinking tokens Gemini should use; -1 uses dynamic thinking, 0 disables thinking, and fixed budgets start at 512 tokens.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Flash Generate content Subscription 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (-1…24576) -1 Number of thinking tokens Gemini should use; 0 disables thinking and -1 uses dynamic thinking.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Pro Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (128…32768) Maximum number of thinking tokens Gemini should use before producing the final answer.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Pro Generate content Subscription 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (128…32768) Maximum number of thinking tokens Gemini should use before producing the final answer.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3 Flash Preview Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3 Flash Preview Generate content Subscription 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "high" Controls Gemini 3 Flash reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Flash Lite Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Flash Lite Preview Generate content Retired 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Flash Lite Preview Generate content Subscription Retired 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "high" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Flash Lite Generate content Subscription 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Pro Preview Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Pro Preview Customtools Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Pro Preview Generate content Subscription 8 params
SDK method models.generateContent Subscription
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (low | high) "high" Controls Gemini 3 Pro reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.5 Flash Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "medium" Controls Gemini 3.5 Flash reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.5 Flash Lite Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "medium" Controls Gemini 3.5 Flash reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.6 Flash Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.7 Flash Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini Flash Latest Generate content 8 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer 0 Number of thinking tokens Gemini should use; -1 uses dynamic thinking, 0 disables thinking, and fixed budgets start at 512 tokens.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemma 3 12B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate; output shares the 128K context window with the input.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 3 1B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…32768) Maximum number of output tokens the model may generate; output shares the 32K context window with the input.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 3 27B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate; output shares the 128K context window with the input.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 3 4B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…131072) Maximum number of output tokens the model may generate; output shares the 128K context window with the input.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 3n E2B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…32768) Maximum number of output tokens the model may generate; output shares the 32K context window with the input.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 3n E4B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…32768) Maximum number of output tokens the model may generate; output shares the 32K context window with the input.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 4 12B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Google's standardized Gemma 4 sampling uses 1.0.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 4 26B A4B IT Generate content 9 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer 0 Number of thinking tokens Gemini should use; -1 uses dynamic thinking, 0 disables thinking, and fixed budgets start at 512 tokens.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | high) Toggles Gemma 4 reasoning; high enables thinking and minimal disables it.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemma 4 31B IT Generate content 9 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (0…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer 0 Number of thinking tokens Gemini should use; -1 uses dynamic thinking, 0 disables thinking, and fixed budgets start at 512 tokens.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | high) Toggles Gemma 4 reasoning; high enables thinking and minimal disables it.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemma 4 E2B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Google's standardized Gemma 4 sampling uses 1.0.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.
Gemma 4 E4B IT Generate content 4 params
SDK method models.generateContent API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Google's standardized Gemma 4 sampling uses 1.0.
Top P
top_p
number (0…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (0…+∞) 64 Limits generation to the selected number of highest-probability tokens.

Z.ai

GLM-4.5 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.5-Air Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.5-Air Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.5-AirX Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.5-Flash Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.5 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.5-X Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…98304) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.6 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.6 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.7 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.7-Flash Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.7-FlashX Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-4.7 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5-Turbo Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5-Turbo Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5.1 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5.1 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5.2 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Reasoning effort
reasoning_effort
enum (none | minimal | low | medium | high | xhigh | max) "max" Controls how much reasoning effort GLM-5.2 spends when thinking is enabled.
Only when thinking.type = "enabled"
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5.2 Chat Completions Subscription 7 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Toggles the model's extended reasoning before it produces the final answer.
Reasoning effort
reasoning_effort
enum (none | minimal | low | medium | high | xhigh | max) "max" Controls how much reasoning effort GLM-5.2 spends when thinking is enabled.
Only when thinking.type = "enabled"
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GLM-5.3 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…131072) 65536 Maximum number of tokens to generate in the response.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Not when do_sample = false
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Not when do_sample = false
Do sample
do_sample
boolean true When false, the model uses greedy decoding and ignores temperature and top_p.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled) "enabled" GLM-5.3 always engages in extended reasoning; thinking cannot be disabled.
Reasoning effort
reasoning_effort
enum (low | high | max) "max" Controls how much reasoning effort GLM-5.3 spends on its always-on thinking.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.

Mistral

Codestral 2508 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Codestral Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Devstral 2512 Chat Completions Retired 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Devstral Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Magistral Medium Latest Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Reasoning 1 param
Parameter Type Default Description Condition
Prompt mode
prompt_mode
enum (reasoning) Enables Mistral's reasoning system prompt; leave unset to disable the default reasoning behavior.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Magistral Small Latest Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Reasoning 1 param
Parameter Type Default Description Condition
Prompt mode
prompt_mode
enum (reasoning) Enables Mistral's reasoning system prompt; leave unset to disable the default reasoning behavior.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Ministral 14B 2512 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Ministral 14B Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Ministral 3B 2512 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Ministral 3B Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Ministral 8B 2512 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Ministral 8B Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Large 2512 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Large Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Medium 3 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Medium 3.5 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Medium Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Small 2603 Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral Small Latest Chat Completions 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.
Mistral NeMo Chat Completions Deprecated 9 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Stop sequence
stop
string Stops generation when this string is detected.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1.5 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Random seed
random_seed
integer (0…+∞) Seed used for deterministic sampling when reproducible outputs are desired.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes repeated words or phrases to encourage a wider variety of generated content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes words based on how often they already appear in the generated text.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns normal text or JSON mode output.
Metadata 1 param
Parameter Type Default Description Condition
Safe prompt
safe_prompt
boolean false Controls whether Mistral injects its safety prompt before the conversation.

NVIDIA

DeepSeek v4 Flash 0731 Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…16384) 4096 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Seed
seed
integer (0…18446744073709552000) 0 Best-effort deterministic sampling seed. Changing the seed produces a different response with similar characteristics. Fix the seed to reproduce results.
GLiNER PII Chat Completions 4 params
SDK method chat.completions.create API key
Sampling 1 param
Parameter Type Default Description Condition
Threshold
threshold
number (0…1) 0.5 Confidence threshold for entity detection. Lower values detect more entities but may include false positives.
Metadata 3 params
Parameter Type Default Description Condition
Chunk length
chunk_length
integer (1…2048) 384 Context window size for processing. Longer texts are automatically split into chunks with overlap for complete coverage. Must be greater than overlap.
Overlap
overlap
integer (0…512) 128 Token overlap between chunks to prevent entity clipping. Must be less than chunk_length.
Flat NER
flat_ner
boolean false When true, prevents overlapping entity spans. When false, may return nested entities such as both a full name and its constituent first name.
Llama 3.1 NemoGuard 8B Topic Control Chat Completions 6 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 1024 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2) 0.5 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Llama 3.1 Nemotron Nano 8B v1 Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…16384) 4096 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Seed
seed
integer (0…18446744073709552000) 0 Best-effort deterministic sampling seed. Changing the seed produces a different response with similar characteristics. Fix the seed to reproduce results.
Llama 3.1 Nemotron Safety Guard 8B v3 Chat Completions 1 param
SDK method chat.completions.create API key
Sampling 1 param
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Llama 3.1 Nemotron Ultra 253B v1 Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…16384) 4096 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Seed
seed
integer (0…18446744073709552000) 0 Best-effort deterministic sampling seed. Changing the seed produces a different response with similar characteristics. Fix the seed to reproduce results.
Llama 3.3 Nemotron Super 49B v1 Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…16384) 4096 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Seed
seed
integer (0…18446744073709552000) 0 Best-effort deterministic sampling seed. Changing the seed produces a different response with similar characteristics. Fix the seed to reproduce results.
Llama 3.3 Nemotron Super 49B v1.5 Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…65536) 65536 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.6 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Seed
seed
integer (0…18446744073709552000) 0 Best-effort deterministic sampling seed. Changing the seed produces a different response with similar characteristics. Fix the seed to reproduce results.
NemoGuard Jailbreak Detect Chat Completions 0 params
SDK method chat.completions.create API key

No parameters documented yet.

Nemotron 3 Nano 30B A3B Chat Completions 5 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…32768) 16384 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (-∞…1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Seed
seed
integer (0…18446744073709552000) Best-effort deterministic sampling seed. Repeated requests with the same seed and parameters should return the same result.
Nemotron 3 Super 120B A12B Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…32768) 16384 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (-∞…1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Seed
seed
integer (0…18446744073709552000) Best-effort deterministic sampling seed. Repeated requests with the same seed and parameters should return the same result.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | high) "high" Controls the reasoning mode. 'none' disables reasoning tokens, 'low' enables low-effort reasoning, and 'high' enables full reasoning.
Reasoning budget
reasoning_budget
integer (-1…32768) 16384 Maximum number of tokens the model may use for internal reasoning before being forced to end the reasoning trace. Use -1 to disable budget enforcement.
Nemotron 3 Ultra 550B A55B Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…32768) 16384 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (-∞…1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Seed
seed
integer (0…18446744073709552000) Best-effort deterministic sampling seed. Repeated requests with the same seed and parameters should return the same result.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | medium | high) "high" Controls the reasoning mode. 'none' disables reasoning tokens, 'medium' enables efficient reasoning, and 'high' enables full reasoning.
Reasoning budget
reasoning_budget
integer (-1…32768) 16384 Maximum number of tokens the model may use for internal reasoning before being forced to end the reasoning trace. Use -1 to disable budget enforcement.
Nemotron 3 Ultra Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…32768) 16384 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (-∞…1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | medium | high) "high" Controls the reasoning mode. 'none' disables reasoning tokens, 'medium' enables efficient reasoning, and 'high' enables full reasoning.
Reasoning budget
reasoning_budget
integer (-1…32768) 16384 Maximum number of tokens the model may use for internal reasoning before being forced to end the reasoning trace. Use -1 to disable budget enforcement.
Nemotron Content Safety Reasoning 4B Chat Completions 5 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…32768) 16384 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (-∞…1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Seed
seed
integer (0…18446744073709552000) Best-effort deterministic sampling seed. Repeated requests with the same seed and parameters should return the same result.
Nemotron Mini 4B Instruct Chat Completions 6 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…4096) 1024 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.2 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.7 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Riva Translate 4B Instruct v1.1 Chat Completions 6 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…4096) 512 Maximum number of tokens to generate. Generation stops when this limit is reached.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 0.9 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Frequency penalty
frequency_penalty
number (-2…2) 0 Penalizes new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Presence penalty
presence_penalty
number (-2…2) 0 Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
USDCode Llama 3.1 70B Instruct Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…2048) 1024 Maximum number of tokens to generate. Generation stops when this limit is reached.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1) 0.1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Not recommended to modify both temperature and top_p in the same call.
Top P
top_p
number (-∞…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability. Not recommended to modify both temperature and top_p in the same call.
Metadata 1 param
Parameter Type Default Description Condition
Expert type
expert_type
enum (auto | code | knowledge | helperfunction) "auto" The type of expert to use. 'knowledge' answers with USD knowledge, 'code' responds with vanilla OpenUSD code, 'helperfunction' uses high-level helper functions, and 'auto' lets the LLM determine which expert to use.

Fireworks

DeepSeek v4 Flash 0731 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
DeepSeek v4 Pro Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
DeepSeek v4 Pro 0813 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
GLM-5p2 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
GPT-OSS 120B Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
GPT-OSS 20B Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Inkling Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Kimi K2p6 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Kimi K2p7 Code Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Kimi K3 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
MiniMax M2p7 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
MiniMax M3 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Muse Glimmer 30B Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Qwen3p7 Plus Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Qwen3p8 2p4t A95B Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.
Qwen3p8 Max Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer (1…100) Limits token sampling to the top K most likely next tokens.
Presence penalty
presence_penalty
number (-2…2 step 0.1) Penalizes tokens that have already appeared to encourage a wider variety of content.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) Penalizes tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) Choose plain text or force valid JSON output.

MiniMax

MiniMax M2 Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2 Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M2.1 Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2.1 Highspeed Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2.1 Highspeed Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M2.1 Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M2.5 Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2.5 Highspeed Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2.5 Highspeed Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M2.5 Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M2.7 Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2.7 Highspeed Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M2.7 Highspeed Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M2.7 Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
MiniMax M3 Chat Completions 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the completion.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Split reasoning
reasoning_split
boolean false Returns the model's reasoning in a separate reasoning_details field instead of inline with the response.
MiniMax M3 Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of tokens to generate in the response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0.01…1 step 0.01) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied. Values must be greater than 0 and at most 1.
Top P
top_p
number (0.01…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.

xAI

Grok 4.20 0309 Non Reasoning Chat Completions 6 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. xAI accepts up to four stop sequences.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.20 0309 Non Reasoning Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. xAI accepts up to four stop sequences.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.20 0309 Reasoning Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.20 0309 Reasoning Chat Completions Subscription 5 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.20 Multi Agent 0309 Responses API 5 params
SDK method responses.create API key
Length 1 param
Parameter Type Default Description Condition
Max output tokens
max_output_tokens
integer (1…+∞) Upper bound for output tokens generated in the Responses API response.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 0.7 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning.effort
enum (low | medium | high | xhigh) Controls whether the Responses API request uses the 4-agent or 16-agent multi-agent setup.
Output 1 param
Parameter Type Default Description Condition
Text format
text.format.type
enum (text | json_object | json_schema) "text" Controls whether the Responses API returns free-form text, JSON mode output, or structured JSON schema output.
Grok 4.3 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high) "low" Controls how much reasoning Grok performs before responding. Set to none for non-reasoning requests.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.3 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high) "low" Controls how much reasoning Grok performs before responding. Set to none for non-reasoning requests.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.5 Chat Completions 6 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "high" Controls how much reasoning Grok performs before responding. Defaults to high. Reasoning cannot be disabled.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.5 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high) "high" Controls how much reasoning Grok performs before responding. Defaults to high. Reasoning cannot be disabled.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.6 Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. xAI accepts up to four stop sequences. grok-4.6 rejects stop because reasoning cannot be disabled.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "high" Controls how much reasoning Grok performs before responding. Defaults to high. Reasoning cannot be disabled. xhigh is grok-4.6 and later.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok 4.6 Chat Completions Subscription 6 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (low | medium | high | xhigh) "high" Controls how much reasoning Grok performs before responding. Defaults to high. Reasoning cannot be disabled. xhigh is grok-4.6 and later.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok Build 0.1 Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.
Grok Build 0.1 Chat Completions Subscription 5 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Upper bound for visible output tokens generated in the chat completion.
Sampling 3 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Seed
seed
integer Optional seed used for decoding when reproducible sampling is desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON mode output, or structured JSON schema output.

Moonshot AI

Kimi K2.5 Chat Completions Deprecated 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none) "auto" Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true
Kimi K2.6 Chat Completions 8 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.
Keep prior reasoning
thinking.keep
enum (all | ) null Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none) "auto" Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true
Kimi K2.6 Chat Completions Subscription 4 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether Kimi reasons step by step before answering. Thinking is enabled by default; set disabled to respond directly.
Keep prior reasoning
thinking.keep
enum (all | ) null Set to all to carry reasoning content from earlier assistant turns into the request; leave null to send only the final answers.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Kimi K2.7 Code Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) 32768 Maximum number of tokens to generate in the chat completion, including reasoning tokens.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled) "enabled" Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none) "auto" Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Forces the response into plain text, a JSON object, or JSON matching a provided schema.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true
Kimi K2.7 Code Highspeed Chat Completions 7 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) 32768 Maximum number of tokens to generate in the chat completion, including reasoning tokens.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled) "enabled" Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none) "auto" Controls tool calling: auto lets the model decide and none blocks tool calls. Moonshot rejects required while thinking is enabled, which is the default on this model.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Forces the response into plain text, a JSON object, or JSON matching a provided schema.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true
Kimi K2.7 Code Highspeed Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion, covering both thinking and the final answer.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled) "enabled" Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Kimi K2.7 Code Chat Completions Subscription 3 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion, covering both thinking and the final answer.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled) "enabled" Thinking is always on for this model. Enabled is the only accepted value; disabled returns an error.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Kimi K3 Chat Completions 6 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) Controls whether Kimi reasons step by step before answering, or responds directly when set to disabled.
Reasoning effort
reasoning_effort
enum (low | high | max) "max" Controls how much reasoning Kimi performs before answering. Thinking is always on for this model, so it cannot be turned off.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) "auto" Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Forces the response into plain text, a JSON object, or JSON matching a provided schema.
Moonshot v1 128K Chat Completions Deprecated 11 params
SDK method chat.completions.create API key
Length 3 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Number of completions
n
integer (1…5) 1 How many chat completion choices to generate for the request.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to talk about new topics.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) "auto" Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true
Moonshot v1 32K Chat Completions Deprecated 11 params
SDK method chat.completions.create API key
Length 3 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Number of completions
n
integer (1…5) 1 How many chat completion choices to generate for the request.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to talk about new topics.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) "auto" Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true
Moonshot v1 8K Chat Completions Deprecated 11 params
SDK method chat.completions.create API key
Length 3 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate in the chat completion.
Stop sequence
stop
string Stops generation when this sequence is produced. Moonshot accepts up to 5 sequences of at most 32 bytes each.
Number of completions
n
integer (1…5) 1 How many chat completion choices to generate for the request.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…1 step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to talk about new topics.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) "auto" Controls tool calling: auto lets the model decide, none blocks tool calls, and required forces one.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Observability 2 params
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Top log probabilities
top_logprobs
integer (0…20) Number of most likely tokens to return a log probability for at each position.
Only when logprobs = true

Vertex

Gemini 2.5 Flash Vertex generate content Deprecated 12 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Flash Lite Vertex generate content Deprecated 10 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 2.5 Pro Vertex generate content Deprecated 12 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3 Flash Preview Vertex generate content 13 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Flash Lite Vertex generate content 13 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Pro Preview Vertex generate content 13 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.1 Pro Preview Customtools Vertex generate content 13 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.5 Flash Vertex generate content 13 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Presence penalty
generationConfig.presencePenalty
number (-2…1) Penalises tokens that already appeared, pushing the model toward new topics.
Frequency penalty
generationConfig.frequencyPenalty
number (-2…1) Penalises tokens in proportion to how often they have already appeared.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "medium" Controls Gemini 3.5 Flash reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.5 Flash Lite Vertex generate content 11 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "medium" Controls Gemini 3.5 Flash reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.6 Flash Vertex generate content 11 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.
Gemini 3.7 Flash Vertex generate content 11 params
SDK method generateContent API key
Length 3 params
Parameter Type Default Description Condition
Max output tokens
generationConfig.maxOutputTokens
integer (1…65536) Maximum number of tokens to include in a response candidate.
Candidate count
generationConfig.candidateCount
integer (1…8) How many independent completions to generate for one request.
Stop sequences
generationConfig.stopSequences
string A list of strings where the model stops generating further tokens.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
generationConfig.temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
generationConfig.topP
number (0…1 step 0.01) 0.95 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
generationConfig.topK
integer (1…+∞) 64 Limits token sampling to the top K most likely next tokens.
Seed
generationConfig.seed
integer Optional seed used for decoding when reproducible sampling is desired.
Reasoning 3 params
Parameter Type Default Description Condition
Thinking budget
generationConfig.thinkingConfig.thinkingBudget
integer (0…+∞) Token budget the model may spend on internal reasoning before answering. 0 disables thinking. Cannot be combined with thinkingLevel.
Not when generationConfig.thinkingConfig.thinkingLevel ≠ null
Include thoughts
generationConfig.thinkingConfig.includeThoughts
boolean false Controls whether Gemini returns available thought summaries in the response parts.
Thinking level
generationConfig.thinkingConfig.thinkingLevel
enum (minimal | low | medium | high) "minimal" Controls Gemini 3.1 Flash-Lite reasoning effort.
Not when generationConfig.thinkingConfig.thinkingBudget ≠ null
Output 1 param
Parameter Type Default Description Condition
Response MIME type
generationConfig.responseMimeType
enum (text/plain | application/json) "text/plain" MIME type for generated text candidates.

Cohere

Command A 03 2025 Cohere Chat 12 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (REQUIRED | NONE) Forces the model to either call a tool or skip tool calls for this request.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command A Plus 05 2026 Cohere Chat 12 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (REQUIRED | NONE) Forces the model to either call a tool or skip tool calls for this request.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command A Reasoning 08 2025 Cohere Chat 14 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "disabled" Controls whether the model reasons step by step before producing its final answer.
Thinking token budget
thinking.token_budget
integer (1…+∞) Maximum number of tokens the model may spend on reasoning before answering.
Only when thinking.type = "enabled"
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (REQUIRED | NONE) Forces the model to either call a tool or skip tool calls for this request.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command A Translate 08 2025 Cohere Chat 12 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (REQUIRED | NONE) Forces the model to either call a tool or skip tool calls for this request.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command A Vision 07 2025 Cohere Chat 12 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (REQUIRED | NONE) Forces the model to either call a tool or skip tool calls for this request.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command R 08 2024 Cohere Chat 11 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT | OFF) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command R Plus 08 2024 Cohere Chat 11 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT | OFF) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.
Command R7B 12 2024 Cohere Chat 12 params
SDK method chat API key
Length 2 params
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop sequences
stop_sequences
string Stops generation when one of these sequences is detected; up to five are allowed.
Sampling 6 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…+∞ step 0.1) 0.3 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
p
number (0.01…0.99 step 0.01) 0.75 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
k
integer (0…500) 0 Limits sampling to the K most likely tokens; 0 disables top-k sampling.
Frequency penalty
frequency_penalty
number (0…1 step 0.1) 0 Penalizes tokens proportional to how often they have already appeared to reduce repetition.
Presence penalty
presence_penalty
number (0…1 step 0.1) 0 Penalizes tokens that have already appeared to encourage a wider variety of content.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (REQUIRED | NONE) Forces the model to either call a tool or skip tool calls for this request.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Controls whether the model returns normal text or JSON object output.
Observability 1 param
Parameter Type Default Description Condition
Log probabilities
logprobs
boolean false Controls whether the response includes log probabilities for the generated tokens.
Metadata 1 param
Parameter Type Default Description Condition
Safety mode
safety_mode
enum (CONTEXTUAL | STRICT) "CONTEXTUAL" Controls Cohere's built-in safety instructions applied to the generation.

Meta

Llama 3.3 70B Instruct Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer Limits generation to the selected number of highest-probability tokens.
Repetition penalty
repetition_penalty
number Penalizes tokens that have already appeared to reduce repetition in the output.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
Llama 3.3 8B Instruct Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer Limits generation to the selected number of highest-probability tokens.
Repetition penalty
repetition_penalty
number Penalizes tokens that have already appeared to reduce repetition in the output.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
Llama 4 Maverick 17B 128E Instruct FP8 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer Limits generation to the selected number of highest-probability tokens.
Repetition penalty
repetition_penalty
number Penalizes tokens that have already appeared to reduce repetition in the output.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
Llama 4 Scout 17B 16E Instruct FP8 Chat Completions 7 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Top K
top_k
integer Limits generation to the selected number of highest-probability tokens.
Repetition penalty
repetition_penalty
number Penalizes tokens that have already appeared to reduce repetition in the output.
Tools 1 param
Parameter Type Default Description Condition
Tool choice
tool_choice
enum (auto | none | required) Controls whether the model may call tools, must call one, or skips tool calls.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_schema) "text" Controls whether the model returns normal text or a schema-constrained JSON object.
Muse Spark 1.1 Chat Completions 9 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness; use this or Top P, but not both.
Not when top_p ≠ 1
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling; use this or Temperature, but not both.
Not when temperature ≠ 1
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens in proportion to how often they have appeared, reducing repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging new topics.
Random seed
seed
integer Requests best-effort deterministic sampling for repeated requests.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high | xhigh) Controls how much reasoning the model should perform before producing an answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON, or schema-constrained JSON.
Metadata 1 param
Parameter Type Default Description Condition
Prompt cache retention
prompt_cache_retention
enum (in_memory | 24h) Controls whether the prompt cache stays in memory or persists for up to 24 hours.
Muse Spark 1.2 Contributor Chat Completions 9 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness; use this or Top P, but not both.
Not when top_p ≠ 1
Top P
top_p
number (0.01…1 step 0.01) 1 Controls nucleus sampling; use this or Temperature, but not both.
Not when temperature ≠ 1
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens in proportion to how often they have appeared, reducing repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging new topics.
Random seed
seed
integer Requests best-effort deterministic sampling for repeated requests.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high | xhigh) Controls how much reasoning the model should perform before producing an answer.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object | json_schema) "text" Controls whether the model returns text, JSON, or schema-constrained JSON.
Metadata 1 param
Parameter Type Default Description Condition
Prompt cache retention
prompt_cache_retention
enum (in_memory | 24h) Controls whether the prompt cache stays in memory or persists for up to 24 hours.

Groq

GPT-OSS 120B Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. Groq accepts up to four stop sequences.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | default) "default" Controls whether the model reasons before answering. 'none' disables reasoning; 'default' lets the model reason.
Reasoning format
reasoning_format
enum (hidden | raw | parsed) Controls how reasoning tokens are returned — hidden from the response, raw within the content, or parsed into a separate field.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GPT-OSS 20B Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. Groq accepts up to four stop sequences.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | default) "default" Controls whether the model reasons before answering. 'none' disables reasoning; 'default' lets the model reason.
Reasoning format
reasoning_format
enum (hidden | raw | parsed) Controls how reasoning tokens are returned — hidden from the response, raw within the content, or parsed into a separate field.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
GPT-OSS Safeguard 20B Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. Groq accepts up to four stop sequences.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | default) "default" Controls whether the model reasons before answering. 'none' disables reasoning; 'default' lets the model reason.
Reasoning format
reasoning_format
enum (hidden | raw | parsed) Controls how reasoning tokens are returned — hidden from the response, raw within the content, or parsed into a separate field.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Qwen3 32B Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. Groq accepts up to four stop sequences.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | default) "default" Controls whether the model reasons before answering. 'none' disables reasoning; 'default' lets the model reason.
Reasoning format
reasoning_format
enum (hidden | raw | parsed) Controls how reasoning tokens are returned — hidden from the response, raw within the content, or parsed into a separate field.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
Qwen3.6 27B Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. Groq accepts up to four stop sequences.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | default) "default" Controls whether the model reasons before answering. 'none' disables reasoning; 'default' lets the model reason.
Reasoning format
reasoning_format
enum (hidden | raw | parsed) Controls how reasoning tokens are returned — hidden from the response, raw within the content, or parsed into a separate field.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.

DeepSeek

DeepSeek Chat Chat Completions Retired 4 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (disabled | enabled) "disabled" Controls whether DeepSeek uses thinking mode before producing the final answer.
DeepSeek Reasoner Chat Completions Retired 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether DeepSeek uses thinking mode before producing the final answer.
Reasoning effort
reasoning_effort
enum (high | max) "high" Controls DeepSeek thinking effort when thinking mode is enabled.
Only when thinking.type = "enabled"
DeepSeek-V4 Flash Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether DeepSeek uses thinking mode before producing the final answer.
Reasoning effort
reasoning_effort
enum (high | max) "high" Controls DeepSeek thinking effort when thinking mode is enabled. For compatibility, low, medium, and xhigh are also accepted, but low and medium are mapped to high, and xhigh is mapped to max.
Only when thinking.type = "enabled"
DeepSeek-V4 Pro Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) 4096 Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 1 Controls nucleus sampling. In DeepSeek thinking mode this parameter is accepted for compatibility but has no effect.
Not when thinking.type = "enabled"
Reasoning 2 params
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether DeepSeek uses thinking mode before producing the final answer.
Reasoning effort
reasoning_effort
enum (high | max) "high" Controls DeepSeek thinking effort when thinking mode is enabled.
Only when thinking.type = "enabled"

Perplexity

Sonar Chat Completions 12 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…128000) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Metadata 9 params
Parameter Type Default Description Condition
Search mode
search_mode
enum (web | academic | sec) Selects the corpus the model searches when grounding its answer.
Search recency filter
search_recency_filter
enum (hour | day | week | month | year) Restricts web search results to a recent time window.
Search domain filter
search_domain_filter
string Limits search to, or excludes, specific domains.
Search after date
search_after_date_filter
string Restricts search results to content published after this date (MM/DD/YYYY).
Search before date
search_before_date_filter
string Restricts search results to content published before this date (MM/DD/YYYY).
Search context size
web_search_options.search_context_size
enum (low | medium | high) "low" Controls how much web search context is retrieved before generating the answer.
Return images
return_images
boolean false Controls whether the response may include related images from the search.
Return related questions
return_related_questions
boolean false Controls whether the response includes suggested follow-up questions.
Disable search
disable_search
boolean false Turns off web search so the model answers from its own knowledge only.
Sonar Deep Research Chat Completions 12 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…128000) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 1 param
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (minimal | low | medium | high) Controls how much reasoning and searching the model performs before producing the report.
Metadata 8 params
Parameter Type Default Description Condition
Search mode
search_mode
enum (web | academic | sec) Selects the corpus the model searches when grounding its answer.
Search recency filter
search_recency_filter
enum (hour | day | week | month | year) Restricts web search results to a recent time window.
Search domain filter
search_domain_filter
string Limits search to, or excludes, specific domains.
Search after date
search_after_date_filter
string Restricts search results to content published after this date (MM/DD/YYYY).
Search before date
search_before_date_filter
string Restricts search results to content published before this date (MM/DD/YYYY).
Search context size
web_search_options.search_context_size
enum (low | medium | high) "low" Controls how much web search context is retrieved before generating the answer.
Return images
return_images
boolean false Controls whether the response may include related images from the search.
Return related questions
return_related_questions
boolean false Controls whether the response includes suggested follow-up questions.
Sonar Pro Chat Completions 12 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…128000) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Metadata 9 params
Parameter Type Default Description Condition
Search mode
search_mode
enum (web | academic | sec) Selects the corpus the model searches when grounding its answer.
Search recency filter
search_recency_filter
enum (hour | day | week | month | year) Restricts web search results to a recent time window.
Search domain filter
search_domain_filter
string Limits search to, or excludes, specific domains.
Search after date
search_after_date_filter
string Restricts search results to content published after this date (MM/DD/YYYY).
Search before date
search_before_date_filter
string Restricts search results to content published before this date (MM/DD/YYYY).
Search context size
web_search_options.search_context_size
enum (low | medium | high) "low" Controls how much web search context is retrieved before generating the answer.
Return images
return_images
boolean false Controls whether the response may include related images from the search.
Return related questions
return_related_questions
boolean false Controls whether the response includes suggested follow-up questions.
Disable search
disable_search
boolean false Turns off web search so the model answers from its own knowledge only.
Sonar Reasoning Pro Chat Completions 12 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…128000) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Metadata 9 params
Parameter Type Default Description Condition
Search mode
search_mode
enum (web | academic | sec) Selects the corpus the model searches when grounding its answer.
Search recency filter
search_recency_filter
enum (hour | day | week | month | year) Restricts web search results to a recent time window.
Search domain filter
search_domain_filter
string Limits search to, or excludes, specific domains.
Search after date
search_after_date_filter
string Restricts search results to content published after this date (MM/DD/YYYY).
Search before date
search_before_date_filter
string Restricts search results to content published before this date (MM/DD/YYYY).
Search context size
web_search_options.search_context_size
enum (low | medium | high) "low" Controls how much web search context is retrieved before generating the answer.
Return images
return_images
boolean false Controls whether the response may include related images from the search.
Return related questions
return_related_questions
boolean false Controls whether the response includes suggested follow-up questions.
Disable search
disable_search
boolean false Turns off web search so the model answers from its own knowledge only.

Xiaomi

MiMo v2.5 Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, covering both the thinking trace and the final answer.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values are more focused; higher values are more varied. Ignored while thinking is enabled, where it is forced to 1.0.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 0.95 Nucleus sampling cutoff. Ignored while thinking is enabled, where it is forced to 0.95.
Not when thinking.type = "enabled"
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether MiMo reasons step by step before answering. Enabled by default; set disabled to respond directly.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
MiMo v2.5 Pro Chat Completions 8 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, covering both the thinking trace and the final answer.
Stop sequences
stop
string Up to a few sequences where generation stops; the stop text is not included in the output.
Sampling 4 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values are more focused; higher values are more varied. Ignored while thinking is enabled, where it is forced to 1.0.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 0.95 Nucleus sampling cutoff. Ignored while thinking is enabled, where it is forced to 0.95.
Not when thinking.type = "enabled"
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens in proportion to how often they have appeared, reducing verbatim repetition.
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether MiMo reasons step by step before answering. Enabled by default; set disabled to respond directly.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.
MiMo v2.5 Chat Completions Subscription 5 params
SDK method chat.completions.create Subscription
Length 1 param
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of tokens to generate, covering both the thinking trace and the final answer.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) 1 Controls randomness. Lower values are more focused; higher values are more varied. Ignored while thinking is enabled, where it is forced to 1.0.
Not when thinking.type = "enabled"
Top P
top_p
number (0…1 step 0.01) 0.95 Nucleus sampling cutoff. Ignored while thinking is enabled, where it is forced to 0.95.
Not when thinking.type = "enabled"
Reasoning 1 param
Parameter Type Default Description Condition
Thinking mode
thinking.type
enum (enabled | disabled) "enabled" Controls whether MiMo reasons step by step before answering. Enabled by default; set disabled to respond directly.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.

Cerebras

Zai GLM-4.7 Chat Completions 10 params
SDK method chat.completions.create API key
Length 2 params
Parameter Type Default Description Condition
Max completion tokens
max_completion_tokens
integer (1…+∞) Maximum number of output tokens the model may generate, including reasoning tokens.
Stop
stop
string A string or list of strings where the API will stop generating further tokens. Cerebras accepts up to four stop sequences.
Sampling 5 params
Parameter Type Default Description Condition
Temperature
temperature
number (0…2 step 0.1) Controls randomness. Lower values make outputs more focused; higher values make them more varied. Adjust this or top_p, not both.
Top P
top_p
number (0…1 step 0.01) Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Frequency penalty
frequency_penalty
number (-2…2 step 0.1) 0 Penalizes tokens by how often they have appeared, reducing verbatim repetition.
Presence penalty
presence_penalty
number (-2…2 step 0.1) 0 Penalizes tokens that have already appeared, encouraging the model to introduce new topics.
Seed
seed
integer Seed used for best-effort deterministic sampling when reproducible outputs are desired.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | low | medium | high) Controls how much reasoning the model performs before answering. 'none' disables reasoning.
Clear thinking
clear_thinking
boolean true When true, the model's thinking from previous turns is excluded from the conversation context; when false, it is preserved, which is useful for agentic workflows.
Output 1 param
Parameter Type Default Description Condition
Response format
response_format.type
enum (text | json_object) "text" Forces the response into plain text or a JSON object.

Thinking Machines

Inkling Chat Completions 5 params
SDK method chat.completions.create API key
Length 1 param
Parameter Type Default Description Condition
Max tokens
max_tokens
integer (1…+∞) Maximum number of output tokens the model may generate.
Sampling 2 params
Parameter Type Default Description Condition
Temperature
temperature
number 1 Controls randomness. Lower values make outputs more focused; higher values make them more varied.
Top P
top_p
number (0…1) 1 Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.
Reasoning 2 params
Parameter Type Default Description Condition
Reasoning effort
reasoning_effort
enum (none | minimal | low | medium | high | xhigh) "high" Controls how much thinking Inkling performs before answering. Accepts a preset name or a number between 0.0 and 0.99; presets map to numeric effort levels (none=0.0, minimal=0.1, low=0.2, medium=0.7, high=0.9, xhigh=0.99).
Separate reasoning
separate_reasoning
boolean true Returns the model's reasoning in a dedicated reasoning_content field instead of interleaving it with the final message content.

Browse by API parameter

All API parameters →

Open any parameter to see its default, range, and every model that accepts it. Counting weights instead? That's a different thing.

temperature 236 top_p 227 max_tokens 175 response_format.type 107 max_completion_tokens 101 thinking.type 74 frequency_penalty 64 presence_penalty 64 top_k 61 inferenceConfig.maxTokens 56 inferenceConfig.stopSequences 56 inferenceConfig.temperature 56 inferenceConfig.topP 56 reasoning_effort 54 stop 51 extra_body.top_k 41 seed 37 generationConfig.maxOutputTokens 33 generationConfig.responseMimeType 33 generationConfig.seed 33 generationConfig.temperature 33 generationConfig.topK 33 generationConfig.topP 33 generationConfig.thinkingConfig.includeThoughts 32 thinking.budget_tokens 26 generationConfig.thinkingConfig.thinkingLevel 23 do_sample 22 thinking.display 22 tool_choice 21 random_seed 20 safe_prompt 20 extra_body.enable_thinking 19 extra_body.thinking_budget 19 generationConfig.thinkingConfig.thinkingBudget 19 extra_body.chat_template_kwargs.enable_thinking 18 output_config.effort 16 logprobs 15 reasoning.effort 15 reasoning.summary 14 text.verbosity 14 additionalModelRequestFields.top_k 11 generationConfig.candidateCount 11 generationConfig.stopSequences 11 generationConfig.frequencyPenalty 8 generationConfig.presencePenalty 8 k 8 p 8 reasoning_split 8 safety_mode 8 stop_sequences 8 additionalModelRequestFields.thinking.budget_tokens 7 additionalModelRequestFields.thinking.type 7 top_logprobs 7 reasoning_format 5 additionalModelRequestFields.inferenceConfig.topK 4 repetition_penalty 4 return_images 4 return_related_questions 4 search_after_date_filter 4 search_before_date_filter 4 search_domain_filter 4 search_mode 4 search_recency_filter 4 web_search_options.search_context_size 4 disable_search 3 n 3 reasoning_budget 3 prompt_cache_retention 2 prompt_mode 2 thinking.keep 2 chunk_length 1 clear_thinking 1 expert_type 1 flat_ner 1 max_output_tokens 1 overlap 1 separate_reasoning 1 text.format.type 1 thinking.token_budget 1 threshold 1

How to use

Building with an AI agent? Hit Copy to grab this whole guide as Markdown and paste it in — or point your agent straight at /llms.txt.

modelparams.dev is an open, community-maintained catalog of model parameters. Each entry shows the knobs you can turn — type, default, range, and the conditions that gate it.

The same model accessed via an API key and via a subscription usually exposes a different set of parameters. We list both as separate entries so the data stays honest.

Catalog API

The full catalog is static JSON, CORS-enabled, served from the edge.

curl https://modelparams.dev/api/v1/models.json

Each entry is keyed by provider/model for API-key variants; subscription variants append -subscription.

If you only need the params for one model contract, use the providerless endpoint. Subscription contracts are model slugs with -subscription.

curl https://modelparams.dev/api/v1/models/openai/gpt-5.5.json
curl https://modelparams.dev/api/v1/models/openai/gpt-5.5-subscription.json

Single model

curl https://modelparams.dev/api/v1/models/anthropic/claude-opus-4-7.json
curl https://modelparams.dev/api/v1/models/anthropic/claude-opus-4-7-subscription.json

JSON Schema

Every entry validates against a JSON Schema you can use in your editor or pipeline.

curl https://modelparams.dev/api/v1/schema.json

Add this header to any YAML you author for autocomplete in VS Code:

# yaml-language-server: $schema=https://modelparams.dev/api/v1/schema.json

Logos

Provider logos are available at /assets/logos/{provider}.svg where {provider} is the provider slug. They use currentColor so they inherit your text color.

curl https://modelparams.dev/assets/logos/anthropic.svg

Logos are sourced from the models.dev repo (MIT) and used under nominative fair use.

Contribute

The data lives in YAML under models/{provider}/{model}-{auth}.yaml in the GitHub repo. Open a PR; CI validates against the schema and rebuilds.

Edit on GitHub MIT licensed