Become a provider

A provider is your account on inference.club paired with one or more agents running on your hardware. The agent is a small program (inference-club-agent) that:

1. Heartbeats into inference.club every 30 seconds, advertising which models it's currently serving.
2. Accepts proxied inference requests from inference.club and forwards them to your local LLM server.

You can be a provider on a workstation under your desk, a homelab box, a rented server — anywhere you can run Docker and reach the internet.

What you need

• Hardware that runs an LLM server. Anything with an OpenAI-compatible HTTP API: vLLM, LM Studio, Ollama (with the OpenAI-compat endpoint enabled), llama.cpp's server, etc.
• An inference.club account and an API key (see Quickstart).
• Docker on the machine that runs the agent.
• Network reachability between inference.club's servers and your agent. On localhost this is automatic. From the public internet you'll need either a publicly reachable hostname (Cloudflare Tunnel, Tailscale Funnel, or a VPS reverse proxy) or to wait for the upcoming push-mode option.

How heartbeats work

Every 30 seconds the agent makes one request:

POST https://api.inference.club/api/inference/agent/heartbeat/
Authorization: Bearer <your-api-key>
Content-Type: application/json

{
  "name": "home-rig",
  "callback_url": "http://192.168.5.173:8002/v1",
  "models": [
    {"name": "qwen3-8b", "context_window": 32768}
  ],
  "health": {"up": true}
}

inference.club upserts a Provider record keyed by (your_user, name) and replaces its model set with what you reported. Inference.club considers the provider online if a heartbeat arrived in the last 60 seconds — one missed beat is fine, two starts to look like a problem.

How proxied requests work

When a consumer calls /v1/chat/completions with "model": "qwen3-8b", inference.club:

1. Authenticates the consumer's API key.
2. Looks up the first online provider on that user's account that serves qwen3-8b.
3. POSTs the request body to <provider.callback_url>/chat/completions.
4. Streams the response back to the consumer untouched.

Your agent is what's on the other end of step 3. It forwards the request to the local LLM server you configured (http://localhost:1234/v1 or wherever).

Next steps

• Run an agent — install the agent, configure it, see the heartbeat land.