OSS AI Models API Documentation

Welcome to the Open Source AI Models Database API! This API provides access to a comprehensive database of 50+ popular open source AI models across various categories.

Base URL

https://ossmodels.dev/api

For local development:

http://localhost:3000/api

Authentication

No authentication required. The API is free and open for public use.

Rate Limiting

Limit: 1000 requests per 15-minute window per IP
Headers: Rate limit information is included in response headers

Response Format

All API responses follow this format:

{
  "success": true,
  "data": { /* response data */ },
  "error": "error message (if success is false)"
}

Endpoints

1. Health Check

GET /api/health

Check API status and get basic information.

Response:

{
  "success": true,
  "data": {
    "status": "healthy",
    "timestamp": "2024-01-15T10:30:00.000Z",
    "version": "1.0.0",
    "totalModels": 50
  }
}

2. Get All Models

GET /api/models

Retrieve all AI models with optional filtering, searching, sorting, and pagination.

Query Parameters:

Parameter	Type	Description	Example
`category`	string	Filter by category	`language`, `vision`, `audio`, `multimodal`, `embedding`
`provider`	string	Filter by provider	`Google`, `Meta`, `OpenAI`
`license`	string	Filter by license	`MIT`, `Apache-2.0`
`search`	string	Search in name, description, tags	`transformer`
`inputModality`	string	Filter by input type	`Text`, `Image`, `Audio`
`outputModality`	string	Filter by output type	`Text`, `Image`, `Audio`
`toolCalling`	string	Filter by tool calling support	`Yes`, `No`, `Limited`
`reasoning`	string	Filter by reasoning capability	`Strong`, `Good`, `Basic`, `None`
`sortBy`	string	Sort field	`name`, `date`, `downloads`, `stars`, `parameters`
`order`	string	Sort order	`asc`, `desc`
`limit`	number	Items per page (max 100)	`20`
`offset`	number	Pagination offset	`0`

Example Request:

GET /api/models?category=language&provider=Meta&limit=10&sortBy=stars&order=desc

Response:

{
  "success": true,
  "data": [
    {
      "name": "LLaMA 2",
      "author": "Meta",
      "provider": "Hugging Face",
      "category": "language",
      "description": "A collection of pretrained and fine-tuned large language models...",
      "plainDescription": "A powerful AI that can have conversations...",
      "useCases": ["Chatbots", "Content Writing", "Code Assistance"],
      "inputModalities": ["Text"],
      "outputModalities": ["Text"],
      "technicalSpecs": {
        "parameters": "7B - 70B",
        "memoryRequired": "14GB - 140GB",
        "hardwareRequirement": "GPU recommended",
        "inferenceSpeed": "Fast",
        "formats": ["PyTorch", "GGUF", "AWQ"],
        "toolCalling": "Limited",
        "reasoning": "Good",
        "inputCost": "Free",
        "outputCost": "Free"
      },
      "license": "Custom",
      "downloads": "2.5M",
      "stars": "45.2K",
      "date": "2023-07-18",
      "tags": ["transformer", "instruction-following", "chat"],
      "githubUrl": "https://github.com/facebookresearch/llama",
      "huggingtfaceUrl": "https://huggingface.co/meta-llama"
    }
  ],
  "pagination": {
    "total": 50,
    "limit": 10,
    "offset": 0,
    "hasMore": true
  },
  "filters": {
    "category": "language",
    "provider": "Meta",
    "sortBy": "stars",
    "order": "desc"
  }
}

3. Get Specific Model

GET /api/models/{identifier}

Retrieve a specific model by name or array index.

Parameters:

identifier: Model name (case-insensitive) or array index

Example Requests:

GET /api/models/llama-2
GET /api/models/0

Response:

{
  "success": true,
  "data": {
    "name": "LLaMA 2",
    "author": "Meta",
    // ... full model data
  }
}

4. Get Filter Options

GET /api/filters

Get all available filter options for the database.

Response:

{
  "success": true,
  "data": {
    "categories": ["language", "vision", "audio", "multimodal", "embedding"],
    "providers": ["Google", "Meta", "OpenAI", "Hugging Face", "Microsoft"],
    "licenses": ["MIT", "Apache-2.0", "GPL-3.0", "Custom"],
    "inputModalities": ["Text", "Image", "Audio", "Video"],
    "outputModalities": ["Text", "Image", "Audio", "Embeddings", "Masks"],
    "toolCalling": ["Yes", "No", "Limited"],
    "reasoning": ["Strong", "Good", "Basic", "None"]
  }
}

5. Get Statistics

GET /api/stats

Get database statistics and insights.

Response:

{
  "success": true,
  "data": {
    "totalModels": 50,
    "categories": {
      "language": 17,
      "vision": 12,
      "audio": 9,
      "multimodal": 8,
      "embedding": 4
    },
    "providers": {
      "Google": 8,
      "Meta": 7,
      "Hugging Face": 15
    },
    "licenses": {
      "MIT": 12,
      "Apache-2.0": 18,
      "GPL-3.0": 5
    },
    "totalDownloads": 50000000,
    "totalStars": 500000
  }
}

6. Vercel AI SDK Compatible Endpoint

POST /api/ai/models

Optimized endpoint for Vercel AI SDK integration.

Request Body:

{
  "query": "language model for coding",
  "filters": {
    "category": "language",
    "toolCalling": "Yes"
  },
  "limit": 5
}

Response:

{
  "success": true,
  "models": [
    {
      "id": "llama-2",
      "name": "LLaMA 2",
      "description": "A powerful AI that can have conversations...",
      "provider": "Hugging Face",
      "category": "language",
      "capabilities": {
        "toolCalling": true,
        "reasoning": true,
        "inputModalities": ["Text"],
        "outputModalities": ["Text"]
      },
      "pricing": {
        "input": "Free",
        "output": "Free"
      },
      "specs": {
        "parameters": "7B - 70B",
        "memory": "14GB - 140GB",
        "speed": "Fast"
      }
    }
  ],
  "total": 5,
  "query": "language model for coding",
  "filters": {
    "category": "language",
    "toolCalling": "Yes"
  }
}

Error Responses

404 Not Found

{
  "success": false,
  "error": "Not found",
  "message": "API endpoint /api/invalid not found"
}

500 Internal Server Error

{
  "success": false,
  "error": "Internal server error",
  "message": "Something went wrong"
}

429 Too Many Requests

{
  "success": false,
  "error": "Too many requests from this IP, please try again later.",
  "retryAfter": "15 minutes"
}

Usage Examples

JavaScript/Node.js

// Get all language models
const response = await fetch('https://ossmodels.dev/api/models?category=language');
const data = await response.json();
console.log(data.data); // Array of language models

// Search for specific models
const searchResponse = await fetch('https://ossmodels.dev/api/models?search=transformer&limit=5');
const searchData = await searchResponse.json();

// Get a specific model
const modelResponse = await fetch('https://ossmodels.dev/api/models/llama-2');
const model = await modelResponse.json();

Python

import requests

# Get all models with filtering
response = requests.get('https://ossmodels.dev/api/models', params={
    'category': 'vision',
    'provider': 'Google',
    'limit': 10
})
data = response.json()

# Search for models
search_response = requests.get('https://ossmodels.dev/api/models', params={
    'search': 'image generation',
    'outputModality': 'Image'
})

cURL

# Get all models
curl "https://ossmodels.dev/api/models"

# Filter by category and provider
curl "https://ossmodels.dev/api/models?category=language&provider=OpenAI"

# Get specific model
curl "https://ossmodels.dev/api/models/whisper"

# Get statistics
curl "https://ossmodels.dev/api/stats"

Integration with Vercel AI SDK

import { createOpenAI } from '@ai-sdk/openai';

// Use the models endpoint to discover available models
async function getAvailableModels() {
  const response = await fetch('https://ossmodels.dev/api/ai/models', {
    method: 'POST',
    headers: {
      'Content-Type': 'application/json',
    },
    body: JSON.stringify({
      query: 'coding assistant',
      filters: {
        category: 'language',
        toolCalling: 'Yes'
      },
      limit: 10
    })
  });
  
  const { models } = await response.json();
  return models;
}

// Example usage
const models = await getAvailableModels();
console.log('Available coding models:', models);

CORS Policy

The API supports CORS for browser-based applications:

Production: https://ossmodels.dev, https://www.ossmodels.dev
Development: All origins allowed
Methods: GET, POST, OPTIONS
Headers: Content-Type, Authorization, X-Requested-With

Deployment

Vercel Deployment

npm install
vercel --prod

Docker Deployment

FROM node:18-alpine
WORKDIR /app
COPY package*.json ./
RUN npm install --production
COPY . .
EXPOSE 3000
CMD ["npm", "start"]

Environment Variables

NODE_ENV=production
PORT=3000

Support

Issues: GitHub Issues
API Status: https://ossmodels.dev/api/health
Documentation: https://ossmodels.dev/docs

License

This API and database are open source under the MIT License.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

OSS AI Models API Documentation

Base URL

Authentication

Rate Limiting

Response Format

Endpoints

1. Health Check

2. Get All Models

3. Get Specific Model

4. Get Filter Options

5. Get Statistics

6. Vercel AI SDK Compatible Endpoint

Error Responses

404 Not Found

500 Internal Server Error

429 Too Many Requests

Usage Examples

JavaScript/Node.js

Python

cURL

Integration with Vercel AI SDK

CORS Policy

Deployment

Vercel Deployment

Docker Deployment

Environment Variables

Support

License

FilesExpand file tree

API.md

Latest commit

History

API.md

File metadata and controls

OSS AI Models API Documentation

Base URL

Authentication

Rate Limiting

Response Format

Endpoints

1. Health Check

2. Get All Models

3. Get Specific Model

4. Get Filter Options

5. Get Statistics

6. Vercel AI SDK Compatible Endpoint

Error Responses

404 Not Found

500 Internal Server Error

429 Too Many Requests

Usage Examples

JavaScript/Node.js

Python

cURL

Integration with Vercel AI SDK

CORS Policy

Deployment

Vercel Deployment

Docker Deployment

Environment Variables

Support

License