++
DOCS · QUICKSTART

Two strings.
Sixty seconds.

Ainfera is OpenAI-compatible. Keep your client, your SDK, your prompts — change the base URL and the model string, and every call routes by outcome.

OPENAI SDK · PY/TSCURLHERMESLANGCHAINPYDANTIC AI
STEP 01Mint a keySign up at app.ainfera.ai and mint one scoped key per agent. Caps ride the key, enforced at the router.POST /v1/keys → ainfera_*
STEP 02Change two strings
base_url = "https://api.ainfera.ai/v1"
model = "ainfera-inference"▍
Any OpenAI-compatible client. Nothing else changes.
STEP 03Watch it routeThe dashboard shows every decision live — model, latency, cost, and the signature that lands it in the public audit log.✓ ROUTED · SIGNED · LOGGED
API REFERENCE
METHODENDPOINTWHAT IT DOES
POST/v1/chat/completionsRoute + respond — OpenAI-compatible
GET/v1/modelsCatalog with pass-through prices
GET/v1/audit/publicThe signed audit log — keyless, verifiable by anyone
POST/v1/keysMint a scoped key per agent
$ curl https://api.ainfera.ai/v1/audit/publicDon't trust — verify. No account needed.
++

Your next deploy could route.