Your router access
Plan and profile
Free plan
Complete your profileOptional: where to find you and what you use AI for. Change your plan there too.
Credits
Paid models draw from this balance as you use tokens. Requests stop when it reaches zero.
0.00
credits remaining
Endpoint
OpenAI-compatible. Point any SDK at this base URL. Quickstart with Python and JavaScript examples
/api/public/v1curl /api/public/v1/chat/completions \
-H "Authorization: Bearer YOUR_ROUTER_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"your-model","messages":[{"role":"user","content":"hi"}]}'No models published yet.
API keys
Your keys are separate from the admin's backend credentials.
A guardrail limits what a key's end users may ask for - for an assistant that should only answer about your site, say. Create and inspect guardrails. The response cache answers a request this key has already made - same messages, same settings, temperature: 0, not streamed - from NYO, for free and in milliseconds, for an hour unless you set nyo.cache.ttl_seconds.
| Name | Prefix | Last used | Status | Guardrail | Cache |
|---|
Built-in observability
Track request volume, success rate, latency, tokens, costs and errors.