Introduction
Rikaii sits between your applications and LLM providers. You call OpenAI-compatible POST /v1/chat/completions; we handle routing, metering, optional bring-your-own-key (BYOK), and billing / spend protection according to your workspace settings.
Who this documentation is for
These docs target server-side integrations—backend services, scripts, and automation. Do not embed API keys in browsers or untrusted clients.
Base URL
Production API host:
https://api.rikaii.com
The primary surface documented here is POST /v1/chat/completions, plus GET /v1/models for the public catalog. Account administration (keys, plans, BYOK secrets, routing rules) happens in the Dashboard—not via additional public REST contracts in this site.
Next steps
- Quickstart — create an API key and send your first request.
- Authentication —
Authorization: Bearerwith your API key. - Chat completions — full request/response, streaming, caching, and rate limits.
- Routing engine — PAYG vs BYOK execution, keywords, heuristics.
- Overage protection — dynamic completion ceilings against runaway spend.
- Macroscopic invoicing — how monthly totals and fees are computed.
- Understanding Cost Avoidance — savings vs the gpt-5.5 baseline in the Dashboard.