// Docs
Documentation
Connect to Compyoot via OpenAI-compatible API in under 5 minutes.
Quickstart
1. Set your base URL
const client = new OpenAI({
apiKey: process.env.COMPYOOT_API_KEY,
baseURL: "https://api.compyoot.com/v1"
});2. Route a task
const response = await client.chat.completions.create({
model: "auto",
messages: [{ role: "user", content: "Extract all prices from this listing" }],
task_hint: "extraction"
});
console.log(response.choices[0].message.content);API Reference
POST
/v1/chat/completionsSend a chat completion. Compyoot classifies and routes to the optimal model.
POST
/v1/feedbackSubmit quality feedback for a completed run.
GET
/v1/modelsList all available models and their current status.
GET
/v1/statsGet router performance statistics.
Available models
32 models across 8 providers.
| Model | Provider | Best for |
|---|---|---|
| cerebras-llama | Cerebras | Fast extraction |
| cerebras-qwen | Cerebras | Reasoning |
| groq-llama | Groq | Low-latency |
| groq-scout | Groq | Analysis |
| samba-llama | SambaNova | Balanced |
| samba-r1 | SambaNova | Chain-of-thought |
| cf-kimi | Fireworks | Analysis |
| fw-gptoss | Fireworks | Generation |
| gemini | Long context + vision | |
| openrouter-qwen36 | OpenRouter | Complex reasoning |
| mistral | Mistral | Fast generation |
| haiku | Anthropic | Fast Claude |
| opus | Anthropic | Premium reasoning |
| zai-flash | ZAI | Speed + quality |