modelparams.dev
API GitHub

Providers

Alibaba Anthropic Cerebras Cohere DeepSeek Google Groq Meta MiniMax Mistral Moonshot AI NVIDIA OpenAI Perplexity Thinking Machines xAI Xiaomi Z.ai

Providers

Alibaba Anthropic Cerebras Cohere DeepSeek Google Groq Meta MiniMax Mistral Moonshot AI NVIDIA OpenAI Perplexity Thinking Machines xAI Xiaomi Z.ai
Groq 10 params

Groq Qwen3 32B parameters

These are the parameters modelparams.dev tracks for Groq Qwen3 32B. Each row gives the type, default, valid range or values, and the conditions that gate it. It's the same data the JSON API serves.

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. —

Groq Qwen3 32B API parameters in brief

Groq Qwen3 32B documents 10 API parameters, grouped by what they control:

Frequently asked questions

How many parameters does Groq Qwen3 32B accept?
Groq Qwen3 32B accepts 10 API parameters: max_completion_tokens, temperature, top_p, frequency_penalty, presence_penalty, seed, and more.
What is the default temperature for Groq Qwen3 32B?
The default temperature for Groq Qwen3 32B is 1, within a valid range of 0 to 2.
What is the default top_p for Groq Qwen3 32B?
The default top_p for Groq Qwen3 32B is 1, within a valid range of 0 to 1.

Resources

All Groq models Glossary Full catalog

Qwen3 32B — JSON

The full model definition as served by the API. Copy it or open the endpoint directly.

/api/v1/models/groq/qwen3-32b.json 
{
  "$schema": "https://modelparams.dev/api/v1/schema.json",
  "provider": "groq",
  "authType": "api_key",
  "model": "qwen3-32b",
  "params": [
    {
      "path": "max_completion_tokens",
      "label": "Max tokens",
      "description": "Maximum number of output tokens the model may generate.",
      "group": "generation_length",
      "type": "integer",
      "range": {
        "min": 1
      }
    },
    {
      "path": "temperature",
      "label": "Temperature",
      "description": "Controls randomness. Lower values make outputs more focused; higher values make them more varied.",
      "group": "sampling",
      "type": "number",
      "default": 1,
      "range": {
        "min": 0,
        "max": 2,
        "step": 0.1
      }
    },
    {
      "path": "top_p",
      "label": "Top P",
      "description": "Controls nucleus sampling by limiting generation to tokens within the selected cumulative probability.",
      "group": "sampling",
      "type": "number",
      "default": 1,
      "range": {
        "min": 0,
        "max": 1,
        "step": 0.01
      }
    },
    {
      "path": "frequency_penalty",
      "label": "Frequency penalty",
      "description": "Penalizes tokens by how often they have appeared, reducing verbatim repetition.",
      "group": "sampling",
      "type": "number",
      "default": 0,
      "range": {
        "min": -2,
        "max": 2,
        "step": 0.1
      }
    },
    {
      "path": "presence_penalty",
      "label": "Presence penalty",
      "description": "Penalizes tokens that have already appeared, encouraging the model to introduce new topics.",
      "group": "sampling",
      "type": "number",
      "default": 0,
      "range": {
        "min": -2,
        "max": 2,
        "step": 0.1
      }
    },
    {
      "path": "seed",
      "label": "Seed",
      "description": "Seed used for best-effort deterministic sampling when reproducible outputs are desired.",
      "group": "sampling",
      "type": "integer"
    },
    {
      "path": "stop",
      "label": "Stop",
      "description": "A string or list of strings where the API will stop generating further tokens. Groq accepts up to four stop sequences.",
      "group": "generation_length",
      "type": "string"
    },
    {
      "path": "reasoning_effort",
      "label": "Reasoning effort",
      "description": "Controls whether the model reasons before answering. 'none' disables reasoning; 'default' lets the model reason.",
      "group": "reasoning",
      "type": "enum",
      "default": "default",
      "values": [
        "none",
        "default"
      ]
    },
    {
      "path": "reasoning_format",
      "label": "Reasoning format",
      "description": "Controls how reasoning tokens are returned — hidden from the response, raw within the content, or parsed into a separate field.",
      "group": "reasoning",
      "type": "enum",
      "values": [
        "hidden",
        "raw",
        "parsed"
      ]
    },
    {
      "path": "response_format.type",
      "label": "Response format",
      "description": "Forces the response into plain text or a JSON object.",
      "group": "output_format",
      "type": "enum",
      "default": "text",
      "values": [
        "text",
        "json_object"
      ]
    }
  ]
}

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/params/gpt-5.5.json
curl https://modelparams.dev/api/v1/params/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