Auto Router

Automatic model selection based on prompt intent

Set model to therouter/auto to delegate model selection to TheRouter.ai. The router evaluates prompt complexity and task type, then forwards to a suitable model.

TypeScript

const response = await fetch('https://api.therouter.ai/v1/chat/completions', {
  method: 'POST',
  headers: {
    Authorization: 'Bearer <THEROUTER_API_KEY>',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    model: 'therouter/auto',
    messages: [
      { role: 'user', content: 'Explain quantum entanglement for a 12-year-old.' },
    ],
  }),
});

const data = await response.json();
console.log('Selected model:', data.model);

Restricting candidate models

Use provider preferences to constrain selection to approved model families.

json

{
  "model": "therouter/auto",
  "messages": [{ "role": "user", "content": "Draft an architecture decision record." }],
  "provider": {
    "only": ["anthropic", "openai"]
  }
}

Pricing

Auto Router has no additional surcharge. Billing uses the standard price of the model selected for the request.