AI Router routes each request to the best model for the job — by quality, cost, speed or your own rules. Bring your own provider keys; we never mark up your token spend. Under €1/month, forever.
Overage beyond your plan: €0.30 / 1,000 calls. You always pay your model providers directly — we never touch or mark up your token spend.
Your provider keys are encrypted with AES-256-GCM and never shown again — we only display a mask like
sk-…AB12. Keys are decrypted in memory only, for your own requests, and can be revoked any time.
We don't store your prompts by default. Accounts support email verification, login lockout, session revocation, and self-serve data export/delete (GDPR).
Point any OpenAI SDK or tool at our endpoint and authenticate with your router token (rt_…) from the dashboard.
Hint the router which lens to optimize for: X-Router-Activity: coding|reasoning|image|video|speed|longcontext|voice|cheap.
| Method | Path | Auth | Purpose |
|---|---|---|---|
| POST | /v1/chat/completions | rt_ token or session | Route a chat request through your stored provider key |
| GET | /api/keys | session | List your encrypted provider keys (masked) |
| POST | /api/keys | session (verified) | Store a provider key (encrypted) |
| POST | /api/keys/:id/test | session (verified) | Validate a key with a live probe |
| GET | /api/usage | session | Your routed-call usage & analytics |
| GET | /api/dataset | public | Live benchmark dataset (8 lenses × 3 categories) |
rt_ token or provider keys. Use HTTPS only. Revoke instantly from the dashboard if a token leaks.