Models

ARouter provides access to hundreds of models from top providers through a single unified API. You can browse models on the ARouter website or use the API directly.

Query Parameters

The Models API supports query parameters to filter results.

`output_modalities`

Filter models by their output capabilities. Accepts a comma-separated list of modalities or "all" to include every model.

Value	Description
`text`	Models that produce text output (default)
`image`	Models that generate images
`audio`	Models that produce audio output
`embeddings`	Embedding models
`all`	Include all models, skip modality filtering

# Default — text models only
curl "https://api.arouter.ai/v1/models" \
  -H "Authorization: Bearer lr_live_xxxx"

# Image generation models only
curl "https://api.arouter.ai/v1/models?output_modalities=image" \
  -H "Authorization: Bearer lr_live_xxxx"

# Text and image models
curl "https://api.arouter.ai/v1/models?output_modalities=text,image" \
  -H "Authorization: Bearer lr_live_xxxx"

# All models regardless of modality
curl "https://api.arouter.ai/v1/models?output_modalities=all" \
  -H "Authorization: Bearer lr_live_xxxx"

`supported_parameters`

Filter models by the API parameters they support. For example, to find models that support tool calling:

curl "https://api.arouter.ai/v1/models?supported_parameters=tools" \
  -H "Authorization: Bearer lr_live_xxxx"

List Models

GET /v1/models

Returns the full list of models available to your API key.

curl https://api.arouter.ai/v1/models \
  -H "Authorization: Bearer lr_live_xxxx"

Response Format

{
  "data": [
    {
      "id": "openai/gpt-5.4",
      "canonical_slug": "openai/gpt-5.4",
      "name": "GPT-5.4",
      "created": 1748000000,
      "description": "OpenAI's flagship multimodal model with state-of-the-art performance.",
      "context_length": 128000,
      "architecture": {
        "input_modalities": ["text", "image"],
        "output_modalities": ["text"],
        "tokenizer": "cl100k_base",
        "instruct_type": "chatml"
      },
      "pricing": {
        "prompt": "0.000005",
        "completion": "0.000015",
        "request": "0",
        "image": "0.00765",
        "web_search": "0",
        "internal_reasoning": "0",
        "input_cache_read": "0.0000025",
        "input_cache_write": "0.000005"
      },
      "top_provider": {
        "context_length": 128000,
        "max_completion_tokens": 16384,
        "is_moderated": true
      },
      "supported_parameters": [
        "tools",
        "tool_choice",
        "max_tokens",
        "temperature",
        "top_p",
        "structured_outputs",
        "response_format",
        "stop",
        "frequency_penalty",
        "presence_penalty",
        "seed"
      ],
      "per_request_limits": null,
      "default_parameters": null,
      "expiration_date": null
    }
  ]
}

Model Object Schema

Each model in the data array contains the following fields:

Field	Type	Description
`id`	`string`	Unique model identifier used in API requests, e.g. `"openai/gpt-5.4"`
`canonical_slug`	`string`	Permanent slug for the model that never changes
`name`	`string`	Human-readable display name
`created`	`number`	Unix timestamp of when the model was added to ARouter
`description`	`string`	Detailed description of the model’s capabilities
`context_length`	`number`	Maximum context window size in tokens
`architecture`	`Architecture`	Technical capabilities object
`pricing`	`Pricing`	Cost structure for using this model (USD per token)
`top_provider`	`TopProvider`	Configuration details for the primary provider
`per_request_limits`	`object \| null`	Rate limiting information (`null` if no limits)
`supported_parameters`	`string[]`	Array of supported API parameters
`default_parameters`	`object \| null`	Default parameter values (`null` if none)
`expiration_date`	`string \| null`	Deprecation date (`null` if not deprecated)

Architecture Object

{
  "input_modalities": string[],  // e.g. ["text", "image"]
  "output_modalities": string[], // e.g. ["text"]
  "tokenizer": string,           // e.g. "cl100k_base"
  "instruct_type": string | null // e.g. "chatml", null if not applicable
}

Pricing Object

All pricing values are in USD per token. A value of "0" means the feature is free.

{
  "prompt": string,              // Cost per input token
  "completion": string,          // Cost per output token
  "request": string,             // Fixed cost per API request
  "image": string,               // Cost per image input
  "web_search": string,          // Cost per web search operation
  "internal_reasoning": string,  // Cost for internal reasoning tokens
  "input_cache_read": string,    // Cost per cached input token read
  "input_cache_write": string    // Cost per cached input token write
}

Top Provider Object

{
  "context_length": number,         // Provider-specific context limit
  "max_completion_tokens": number,  // Maximum tokens in response
  "is_moderated": boolean           // Whether content moderation is applied
}

Supported Parameters

The supported_parameters array lists which OpenAI-compatible parameters work with a model:

Parameter	Description
`tools`	Function calling capabilities
`tool_choice`	Tool selection control
`max_tokens`	Response length limiting
`temperature`	Randomness control
`top_p`	Nucleus sampling
`reasoning`	Internal reasoning mode
`include_reasoning`	Include reasoning in response
`structured_outputs`	JSON schema enforcement
`response_format`	Output format specification
`stop`	Custom stop sequences
`frequency_penalty`	Repetition reduction
`presence_penalty`	Topic diversity
`seed`	Deterministic outputs

Using Models

Use the id directly as the model field in your requests:

Python
TypeScript
cURL

from openai import OpenAI

client = OpenAI(
    base_url="https://api.arouter.ai/v1",
    api_key="lr_live_xxxx",
)

# List available models
models = client.models.list()
for model in models.data:
    print(model.id)

# Use a specific model
response = client.chat.completions.create(
    model="anthropic/claude-sonnet-4.6",
    messages=[{"role": "user", "content": "Hello!"}],
)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.arouter.ai/v1",
  apiKey: "lr_live_xxxx",
});

// List available models
const models = await client.models.list();
for (const model of models.data) {
  console.log(model.id);
}

// Use a specific model
const response = await client.chat.completions.create({
  model: "anthropic/claude-sonnet-4.6",
  messages: [{ role: "user", content: "Hello!" }],
});

curl https://api.arouter.ai/v1/models \
  -H "Authorization: Bearer lr_live_xxxx"

Filtering by Supported Parameters

Find models that support tool calling:

curl "https://api.arouter.ai/v1/models?supported_parameters=tools" \
  -H "Authorization: Bearer lr_live_xxxx"

Find models that support structured outputs:

curl "https://api.arouter.ai/v1/models?supported_parameters=structured_outputs" \
  -H "Authorization: Bearer lr_live_xxxx"

Auto Routing

In addition to specific model IDs, ARouter supports automatic model selection:

Model	Description
`"auto"`	ARouter automatically selects the best available model for your request

curl https://api.arouter.ai/v1/chat/completions \
  -H "Authorization: Bearer lr_live_xxxx" \
  -H "Content-Type: application/json" \
  -d '{"model": "auto", "messages": [{"role": "user", "content": "Hello!"}]}'

The response model field always shows the model that was actually used. See Model Routing for details.

Model Variants

You can append suffixes to any model ID to influence routing behavior:

Suffix	Effect
`:nitro`	Route to the highest-throughput instance — optimized for speed
`:floor`	Route to the lowest-cost instance — optimized for price
`:free`	Route to the free-tier instance (rate limits apply)
`:thinking`	Enable extended reasoning / chain-of-thought mode

{"model": "openai/gpt-5.4:nitro"}   // fastest
{"model": "openai/gpt-5.4:floor"}   // cheapest
{"model": "deepseek/deepseek-r1:thinking"} // reasoning mode

See Model Variants for the full reference.

Tokenization

Different models tokenize text differently. Some models (GPT, Claude, Llama) break text into multi-character chunks; others tokenize by character (PaLM). This means token counts — and therefore costs — vary between models even for identical inputs and outputs. Costs are billed according to the tokenizer for the model in use. Use the usage field in each response to get the exact token counts:

{
  "usage": {
    "prompt_tokens": 42,
    "completion_tokens": 128,
    "total_tokens": 170
  }
}

Notes

The model list is filtered by your account’s enabled providers. If a provider is not enabled, its models will not appear.
New models are added automatically as providers release them.
Use model IDs from this list directly in the model field of your chat completion requests.

See Providers for a curated list of available providers and their flagship models.

Get Started

Core Concepts

Routing

Features

Guides

Privacy

Administration

Best Practices

Frameworks & Integrations

For Providers

Support

Query Parameters

`output_modalities`

`supported_parameters`

List Models

Response Format

Model Object Schema

Architecture Object

Pricing Object

Top Provider Object

Supported Parameters

Using Models

Filtering by Supported Parameters

Auto Routing

Model Variants

Tokenization

Notes

Get Started

Core Concepts

Routing

Features

Guides

Privacy

Administration

Best Practices

Frameworks & Integrations

For Providers

Support

​Query Parameters

​output_modalities

​supported_parameters

​List Models

​Response Format

​Model Object Schema

​Architecture Object

​Pricing Object

​Top Provider Object

​Supported Parameters

​Using Models

​Filtering by Supported Parameters

​Auto Routing

​Model Variants

​Tokenization

​Notes

Query Parameters

`output_modalities`

`supported_parameters`

List Models

Response Format

Model Object Schema

Architecture Object

Pricing Object

Top Provider Object

Supported Parameters

Using Models

Filtering by Supported Parameters

Auto Routing

Model Variants

Tokenization

Notes