Sign inGet started

Docs

Quick start, editor setup and FAQ for the OpenAI-compatible API.

Usage & Logs

Usage

Usage in the sidebar. Filter by range (24 h / 7 d / 30 d), workspace, key and model. You see:

  • Spend per day (IDR), requests, tokens in/out, fallback rate.
  • Spend by model and by key.
  • A request table with served model, tokens, latency, status and cost.

Admins and Members see only their workspace; Members see only keys they own or that are shared.

Logs

Logs lists every request in scope with its x-libra-request-id, status, served model and whether a fallback happened. Prompt and completion bodies are shown only for keys whose guardrails have Store request bodies on, and are kept 7 days.

Quote the request id when asking for help.

Response headers

Every API response carries:

  • x-libra-request-id: the id shown in Logs.
  • x-libra-served-by: the concrete provider/model that answered.
  • x-libra-fallback: true if the first choice was unavailable.