# OauthRouter — Complete API Documentation

> OpenAI-compatible LLM router. Users bring their own API keys or OAuth tokens for any of 11+ providers and get a single unified endpoint. No token markup.

## What is OauthRouter

OauthRouter lets developers route LLM requests to multiple AI providers (Anthropic, OpenAI, Google, Mistral, Groq, xAI, DeepSeek, Together, Cohere, Cloudflare AI, OpenRouter) through one OpenAI-compatible API. Unlike OpenRouter, users connect their OWN provider credentials (API keys or OAuth tokens like Claude Code sk-ant-oat) — so you pay providers directly with zero middleman markup.

Key features:
- OpenAI-compatible: drop-in replacement for any `openai.chat.completions.create` call
- OAuth support for Anthropic Claude subscriptions (Claude Code CLI tokens)
- Routing strategies: latest, random, round-robin, failover, least-used
- Model filtering per token based on configured providers
- Message limits by plan (100/mo free, 1k/mo starter, 5k/mo pro, unlimited enterprise)

## Base URLs

- API Base: `https://api.oauthrouter.com/v1` (use this as base_url for SDKs)
- Full chat endpoint: `https://api.oauthrouter.com/v1/chat/completions`
- Models endpoint: `https://api.oauthrouter.com/v1/models`
- Dashboard: `https://oauthrouter.com/dashboard`
- OpenAPI spec: `https://oauthrouter.com/openapi.json`

## Authentication

All API requests require a Bearer token in the Authorization header. Tokens start with `lr_live_` and are unique per account.

```
Authorization: Bearer lr_live_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
```

To get a token:
1. Sign up at https://oauthrouter.com (free tier: 100 messages/month)
2. Add at least one provider in the Providers tab (paste your Anthropic/OpenAI/Google API key)
3. Copy the token from the "My API" tab

## Endpoint: POST /v1/chat/completions

OpenAI-compatible chat completions. Works with the official OpenAI SDKs (Python, Node.js, etc) by setting `base_url` to `https://api.oauthrouter.com/v1`.

### Request body

```json
{
  "model": "anthropic/claude-sonnet-4-5",
  "messages": [
    {"role": "system", "content": "You are helpful."},
    {"role": "user", "content": "Hello!"}
  ],
  "max_tokens": 1024,
  "temperature": 1,
  "top_p": 1,
  "stream": false
}
```

### Parameters

| Name | Type | Required | Default | Description |
|------|------|----------|---------|-------------|
| model | string | yes | — | Model identifier. Format: `provider/model-name` (e.g. `anthropic/claude-sonnet-4-5`) or just `model-name`. |
| messages | array | yes | — | Array of message objects with `role` (system/user/assistant) and `content`. |
| max_tokens | number | no | 8192 | Maximum tokens in the response. |
| temperature | number | no | 1.0 | Sampling temperature (0 to 2). Higher = more creative. |
| top_p | number | no | 1.0 | Nucleus sampling threshold. Use either temperature or top_p. |
| stream | boolean | no | false | If true, returns server-sent events. |

### Response

Standard OpenAI format:

```json
{
  "id": "chatcmpl-xxx",
  "object": "chat.completion",
  "created": 1234567890,
  "model": "claude-sonnet-4-5-20250929",
  "choices": [{
    "index": 0,
    "message": {"role": "assistant", "content": "Hello!"},
    "finish_reason": "stop"
  }],
  "usage": {"prompt_tokens": 10, "completion_tokens": 5, "total_tokens": 15}
}
```

## Endpoint: GET /v1/models

Returns models available to the caller's token. The response depends on whether you're authenticated:

- **Authenticated** (`Authorization: Bearer lr_live_...`): returns only models from providers the user has enabled.
- **No auth**: returns the generic catalog (useful for tooling discovery).

Response format is OpenAI-compatible:

```json
{
  "data": [
    {
      "id": "anthropic/claude-sonnet-4-5",
      "name": "Claude Sonnet 4.5",
      "context_length": 200000,
      "pricing": {"prompt": "3", "completion": "15"}
    }
  ]
}
```

## Supported providers and model IDs

| Provider | Example model IDs |
|----------|-------------------|
| anthropic | claude-opus-4-6, claude-sonnet-4-5, claude-haiku-4-5 |
| openai | gpt-4o, gpt-4o-mini, o1, o1-mini |
| google | gemini-2.5-pro, gemini-2.5-flash, gemini-2.0-flash |
| mistral | mistral-large-latest, mistral-medium-latest, codestral-latest |
| groq | llama-3.3-70b-versatile, llama-3.1-8b-instant, gemma2-9b-it |
| xai | grok-3, grok-3-mini |
| deepseek | deepseek-chat, deepseek-reasoner |
| together | meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo, Qwen/Qwen2.5-72B-Instruct-Turbo |
| cohere | command-r-plus, command-r |
| cloudflare-ai | @cf/meta/llama-3.1-8b-instruct (free tier) |
| openrouter | sk-or-style routing to any OpenRouter model |

Use model IDs as `provider/model-name` (e.g. `anthropic/claude-sonnet-4-5`).

## Error responses

All errors follow a consistent JSON format:

```json
{
  "error": {
    "message": "Invalid or disabled token",
    "type": "authentication_error",
    "docs_url": "https://oauthrouter.com/docs#errors"
  }
}
```

### HTTP status codes

| Code | Type | Cause | Solution |
|------|------|-------|----------|
| 401 | authentication_error | Invalid or missing token | Reveal/regenerate your token at /dashboard/api |
| 400 | invalid_request_error | Malformed request body | Check model name and messages array |
| 429 | rate_limit_error | Monthly message limit or budget exceeded | Upgrade plan or wait for reset |
| 502 | api_error | Upstream provider request failed | Verify your provider API key is valid |

## Rate limits (by plan)

| Plan | Price | Messages/month |
|------|-------|----------------|
| Free | $0 | 100 |
| Starter | $29 | 1,000 |
| Pro | $49 | 5,000 |
| Enterprise | $69 | Unlimited |

Limits reset on the 1st of each month at 00:00 UTC. Bonus messages from referrals add to the monthly limit.

## Code examples

### cURL

```bash
curl https://api.oauthrouter.com/v1/chat/completions \
  -H "Authorization: Bearer lr_live_YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"model":"anthropic/claude-sonnet-4-5","messages":[{"role":"user","content":"Hello!"}]}'
```

### Python (OpenAI SDK)

```python
from openai import OpenAI
client = OpenAI(api_key="lr_live_YOUR_TOKEN", base_url="https://api.oauthrouter.com/v1")
resp = client.chat.completions.create(
    model="anthropic/claude-sonnet-4-5",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(resp.choices[0].message.content)
```

### Node.js / TypeScript (OpenAI SDK)

```javascript
import OpenAI from 'openai';
const client = new OpenAI({
  apiKey: 'lr_live_YOUR_TOKEN',
  baseURL: 'https://api.oauthrouter.com/v1'
});
const resp = await client.chat.completions.create({
  model: 'anthropic/claude-sonnet-4-5',
  messages: [{ role: 'user', content: 'Hello!' }]
});
console.log(resp.choices[0].message.content);
```

### LiteLLM

```python
import litellm
response = litellm.completion(
    model="openai/anthropic/claude-sonnet-4-5",
    api_base="https://api.oauthrouter.com/v1",
    api_key="lr_live_YOUR_TOKEN",
    messages=[{"role": "user", "content": "Hello!"}]
)
```

### LangChain

```python
from langchain_openai import ChatOpenAI
llm = ChatOpenAI(
    base_url="https://api.oauthrouter.com/v1",
    api_key="lr_live_YOUR_TOKEN",
    model="anthropic/claude-sonnet-4-5"
)
```

## Integrations with popular tools

For GUI tools like OpenClaw, Cursor, Cline, Continue.dev, OpenWebUI, LibreChat:

- Provider type: **OpenAI-compatible** or **Custom OpenAI**
- Base URL: **https://api.oauthrouter.com/v1**
- API Key: **lr_live_YOUR_TOKEN**
- Model: **anthropic/claude-sonnet-4-5** (or any supported model ID)

## Common pitfalls

- **Wrong Base URL**: use `api.oauthrouter.com` (with `api.` subdomain), not `oauthrouter.com`.
- **Trailing slash**: some SDKs break if Base URL ends with `/`. Use `/v1` without trailing slash.
- **Slow models timeout**: Claude Opus 4.6 can take 60+ seconds. Set client timeout to 120s+.
- **Model format**: both `provider/model-name` and plain `model-name` work.

## Support

- Dashboard: https://oauthrouter.com/dashboard
- OpenAPI: https://oauthrouter.com/openapi.json
- Docs HTML: https://oauthrouter.com/docs