Quickstart
AsteroidRouter is a unified OpenAI-compatible endpoint for top AI models (GPT, Claude, Gemini, DeepSeek, and more) at low per-token prices. Integrate once and automatically get redundancy, lowest latencies, and real-time fallbacks.
Get an API Key
Before using AsteroidRouter, you must generate an API key to authenticate your HTTP requests.
- Go to the AsteroidRouter Console and log in or register.
- Navigate to the API Keys section from the sidebar navigation.
- Click the Generate API Key button to create a new token.
- Copy the token values immediately (starts with sk-ar-...). For security reasons, the console will not show this token again.
Keep your API key secure
Do not commit your keys to GitHub, embed them inside client-side JS scripts, or share them in public spaces. Always load keys via backend environments or protected vault secrets.
Base URL Configuration
To route requests through AsteroidRouter, redirect the baseURL parameter in your OpenAI SDK (or custom client) to our endpoint:
Verify The Connection
You can verify your connection to the AsteroidRouter infrastructure instantly by firing a basic curl chat-completions request from your terminal:
curl https://inference-api.asteroidrouter.ai/v1/chat/completions \ -H "Content-Type: application/json" \ -H "Authorization: Bearer $ASTEROIDROUTER_API_KEY" \ -d '{ "model": "gpt-5", "messages": [{"role": "user", "content": "Ping"}] }'
A successful connection returns a standard OpenAI-formatted JSON object containing the response text:
{
"id": "chatcmpl-ast888",
"object": "chat.completion",
"created": 1785502900,
"model": "gpt-5",
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": "Pong!"
},
"finish_reason": "stop"
}
],
"usage": {
"prompt_tokens": 8,
"completion_tokens": 2,
"total_tokens": 10
}
}
Fast Integration Tip
Because AsteroidRouter uses a strictly OpenAI-compliant endpoint structure, you can drop-in our configuration to existing open-source web UIs like Chatbox or LibreChat in just 2 minutes!
Supported Models
Route requests to all major frontier and custom models under a single unified pricing agreement:
| Provider | Model Name | Context | Input / 1M | Output / 1M |
|---|---|---|---|---|
| Anthropic | claude-3-5-sonnet | 200k | $3.00 | $15.00 |
| OpenAI | gpt-5 | 128k | $2.50 | $10.00 |
| gemini-1.5-pro | 2M | $1.25 | $5.00 | |
| DeepSeek | deepseek-v3 | 64k | $0.14 | $0.28 |
| xAI | grok-2 | 128k | $2.00 | $10.00 |
Supported Tools
CLI Interface
Control routing groups, run tests, and check server status locally.
Agent SDKs
Link CrewAI, LangChain, and other libraries with instant failover backups.
VS Code extension
Run local autocomplete diagnostics directly inside your favorite IDE editor.
API Endpoint
OpenAI SDK compatible routing logic integrated directly into your backend code.