Models

DocsRouter provides access to 100+ vision-capable models through a unified API. Models are dynamically fetched from OpenRouter, ensuring you always have access to the latest vision models.

View all available models on our Providers page with live filtering and search.

List Models

GET https://api.docsrouter.com/v1/models

Returns all available vision models (models with image input capability) with OpenRouter-compatible schema plus DocsRouter extensions.

Authentication

This endpoint supports both authenticated and unauthenticated requests:

Without authentication: Returns public model information
With authentication: Returns additional pricing details specific to your account tier

# Unauthenticated (public info only)
curl https://api.docsrouter.com/v1/models

# Authenticated (full details)
curl https://api.docsrouter.com/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Query Parameters

Parameter	Type	Default	Description
`provider`	string	all	Filter by provider (e.g., `google`, `openai`, `anthropic`)
`capability`	string	all	Filter by capability: `vision`, `pdf`, `handwriting`
`recommended`	boolean	false	Only return recommended models
`limit`	number	100	Maximum number of models to return

Response

{
  "object": "list",
  "data": [
    {
      "id": "google/gemini-2.0-flash-001",
      "object": "model",
      "created": 1702425600,
      "owned_by": "google",
      "name": "Gemini 2.0 Flash",
      "description": "Fast, cheap, and reliable. Best default choice for most OCR tasks.",
      "context_length": 1048576,
      "architecture": {
        "input_modalities": ["text", "image"],
        "output_modalities": ["text"],
        "tokenizer": "gemini",
        "instruct_type": "gemini"
      },
      "pricing": {
        "prompt": "0.00000010",
        "completion": "0.00000040",
        "image": "0.0001315",
        "request": "0"
      },
      "top_provider": {
        "context_length": 1048576,
        "max_completion_tokens": 8192,
        "is_moderated": false
      },
      "docsrouter": {
        "category": "vision_llm",
        "recommended": true,
        "supports_tables": true,
        "supports_handwriting": true,
        "supports_forms": true,
        "best_for": ["general documents", "invoices", "receipts", "forms"],
        "cost_per_1m_tokens": 0.10
      }
    }
  ]
}

Get Specific Model

GET https://api.docsrouter.com/v1/models/{model_id}

Returns details for a specific model including pricing, capabilities, and context limits.

Path Parameters

Parameter	Description
`model_id`	The model ID (e.g., `google/gemini-2.0-flash-001`)

Example

curl https://api.docsrouter.com/v1/models/google/gemini-2.0-flash-001 \
  -H "Authorization: Bearer YOUR_API_KEY"

Response

{
  "id": "google/gemini-2.0-flash-001",
  "object": "model",
  "created": 1702425600,
  "owned_by": "google",
  "name": "Gemini 2.0 Flash",
  "description": "Fast, cheap, and reliable. Best default choice for most OCR tasks.",
  "context_length": 1048576,
  "architecture": {
    "input_modalities": ["text", "image"],
    "output_modalities": ["text"],
    "tokenizer": "gemini"
  },
  "pricing": {
    "prompt": "0.00000010",
    "completion": "0.00000040",
    "image": "0.0001315"
  },
  "docsrouter": {
    "category": "vision_llm",
    "recommended": true,
    "supports_tables": true,
    "supports_handwriting": true,
    "best_for": ["general documents", "invoices", "receipts"]
  }
}

DocsRouter Extensions

The docsrouter object in each model contains OCR-specific metadata:

Field	Type	Description
`category`	string	Model type: `vision_llm`, `traditional_ocr`, or `specialized_ocr`
`recommended`	boolean	Whether this model is recommended for general use
`supports_tables`	boolean	Can extract tables as structured data
`supports_handwriting`	boolean	Good at recognizing handwritten text
`supports_forms`	boolean	Can extract form key-value pairs
`best_for`	array	List of use cases this model excels at
`cost_per_1m_tokens`	number	Cost per million tokens (USD)

Model Reference

ID	Provider	Cost / 1M Tokens	Best For
`google/gemini-2.0-flash-001`	Google	$0.10	Recommended. Best balance of speed, cost, and accuracy.
`google/gemini-2.5-flash-preview`	Google	$0.15	Complex layouts, multi-column documents.
`google/gemini-2.5-pro-preview`	Google	$2.50	Legal documents, contracts, scientific papers.
`openai/gpt-4o-mini`	OpenAI	$0.15	Simple documents, printed text.
`openai/gpt-4o`	OpenAI	$5.00	Structured extraction, complex reasoning.
`anthropic/claude-sonnet-4`	Anthropic	$3.00	Document analysis, summarization.
`anthropic/claude-3.5-sonnet`	Anthropic	$3.00	Handwriting, dense technical documents.
`mistralai/pixtral-large-2411`	Mistral	$2.00	European documents, multi-language.

Selecting a Model

In Chat Completions (Recommended)

const response = await client.chat.completions.create({
  model: 'google/gemini-2.0-flash-001',
  messages: [...]
});

In Native OCR Endpoint

{
  "model": "anthropic/claude-3.5-sonnet",
  "url": "https://example.com/document.png"
}

Default Model

If no model is specified, google/gemini-2.0-flash-001 is used. This model offers the best balance of:

Speed: ~1-2 seconds per page
Cost: $0.10 per million tokens
Accuracy: 95%+ on printed documents

Model Categories

Vision LLMs

General-purpose models that can understand images and extract text. Best for most use cases.

Traditional OCR (Coming Soon)

Specialized OCR engines like Google Cloud Vision, AWS Textract, and Azure Document Intelligence. Better for high-volume, simple documents.

Specialized OCR (Coming Soon)

Domain-specific models for invoices, receipts, IDs, and other document types.

Additional Endpoints

Get Model Statistics

GET https://api.docsrouter.com/v1/models/stats

Returns aggregate statistics about available models:

{
  "total": 127,
  "by_provider": {
    "google": 15,
    "openai": 8,
    "anthropic": 6,
    "meta-llama": 12,
    ...
  },
  "by_category": {
    "vision_llm": 127
  }
}

Get Recommended Models

GET https://api.docsrouter.com/v1/models/recommended

Returns only the models we recommend for OCR tasks based on quality and cost-effectiveness.

Get Models by Category

GET https://api.docsrouter.com/v1/models/categories

Returns models grouped by category (vision_llm, traditional_ocr, specialized_ocr).

Models

On this page