LLMRPM

LLMRPM documentation

LLMRPM is a single gateway API that sits in front of every model your team already builds against. Authenticate once, keep your existing OpenAI or Anthropic client code, and let LLMRPM handle routing, metering, and clear request windows.

What LLMRPM does

Instead of holding separate accounts and keys for every model provider, you hold one LLMRPM key. Requests you send to LLMRPM are authenticated, checked against your plan's 5-hour, 24-hour, and 7-day request windows, and forwarded to the model you asked for. The response comes back in the exact shape your client expects — no translation layer for you to maintain.

  • Two request formats. An OpenAI-compatible /chat/completions endpoint and an Anthropic-compatible /messages endpoint, so you can keep whichever SDK your codebase already uses.
  • One key, every model. Switch models by changing a string in the request body — see the full catalog on /models.
  • Predictable limits. Fixed 5-hour, 24-hour, and 7-day request windows, all queryable from the API and visible live in your dashboard.

Base URL

Every endpoint documented here is relative to the versioned base URL below. OpenAI SDKs and raw HTTP clients use it directly; Anthropic SDKs need a slightly different base — see Messages for that detail.

https://llmrpm.com/api/v1

Quick start

  1. Create an account and buy a plan. Every paid plan grants 5-hour, 24-hour, and 7-day request allowances — see pricing for the available tiers.
  2. Generate an API key. Create one from /dashboard/api-keys — see Authentication for the full flow.
  3. Make a request. Send it to /api/v1/chat/completions with your key as a bearer token:
cURL
curl https://llmrpm.com/api/v1/chat/completions \
  -H "Authorization: Bearer $LLMRPM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-5",
    "messages": [{ "role": "user", "content": "Say hello in one sentence." }]
  }'

Where to next

Next Authentication