LLMRPM

API endpoints

LLMRPM exposes four endpoints. Two forward model requests, one lists available models, and one reports current usage across your plan's 5-hour, 24-hour, and 7-day windows.

Endpoint reference

All paths are relative to https://llmrpm.com/api/v1 and require the bearer token described in Authentication.

Method & pathDescription
POST/chat/completions
OpenAI-format chat completions. Accepts the standard OpenAI request body, including stream. See Chat completions.
POST/messages
Anthropic-format Messages endpoint. See Messages for the header and auth details specific to Anthropic SDKs.
GET/models
List every model available to your key. See Models.
GET/usage
Current plan, each request window, requests used, and when each window resets. See Usage & limits.

Common behavior

  • Authentication first. Every endpoint checks your bearer token before doing anything else, and returns 401 if it's missing or invalid.
  • Quota next. /chat/completions and /messages check all applicable plan windows before forwarding. /models and /usage do not consume plan requests.
  • JSON in, JSON (or a stream) out. Every endpoint accepts and returns JSON, except a streaming /chat/completions request, which returns server-sent events.
PreviousAuthenticationNext Chat completions