Why route through HYRE
One key, 20+ models
Open-weight models (DeepSeek, GLM, Llama, Qwen, Kimi, MiniMax) plus frontier models (Claude, GPT, Gemini, Grok) behind a single endpoint.
Drop-in OpenAI compatibility
/chat/completions and /models follow the OpenAI schema — the OpenAI SDK, LangChain, and the Vercel AI SDK work unchanged.Crypto-native billing
Prepaid USDC credits, per-call x402 micropayments, or partner invoicing. No card on file, no subscription.
Streaming + tool calls
SSE streaming with usage reporting, function calling (
tools / tool_choice), and response_format all pass through.How access works
Request a key
Tell us who you are and what you’re building — see Get Access. Access is selective; we issue keys manually.
Receive your key
You get a
hyre_gw_* key with agreed daily caps and billing terms (prepaid credits or partner invoicing).Point your app at the gateway
Set the base URL and key in your OpenAI client. Done — see the Quickstart.
Endpoints
| Method | Path | Auth | Description |
|---|---|---|---|
POST | /api/inference/v1/chat/completions | API key or x402 | OpenAI-compatible chat completions (streaming + non-streaming) |
GET | /api/inference/v1/models | none | Public model catalog with live pricing and free-tier flags |
No API key? The gateway also accepts per-call x402 payments (Solana USDC or BSC USD1/USDT/USDC) — a
402 challenge is returned and your wallet or agent pays per request. Keys are the right choice for production integrations; x402 is great for agents paying as they go.