Introduction
Qufas is an OpenAI-compatible multi-model API platform and AI Gateway.
Overview
One API for supported AI models
Qufas provides an OpenAI-compatible developer experience with unified authentication, prepaid billing, observability, model access, and provider execution.
What Qufas is
Applications send documented OpenAI-compatible requests to Qufas. The gateway validates the request, authenticates its API key, resolves an enabled model and provider, records usage, and returns an OpenAI-compatible response.
Compatibility applies to the documented Chat Completions parameter set. Qufas does not currently expose every OpenAI API or parameter.
OpenAI-compatible development
Use the Chat Completions interface for the documented Qufas parameter set.
One model API
Call supported language and image models through one authenticated API.
Server-side API keys
Create and scope qf_sk_ keys in the dashboard and keep secrets on your server.
Usage and observability
Correlate usage, cost, latency, provider attempts, and request IDs.
Stable capabilities
Embeddings, Responses API, public audio/video APIs, reranking, BYOK, guardrails, and smart routing are not part of the stable public API yet.
Model Hub and callable models
The Model Hub is a discovery surface. The authoritative public callable-model contract is GET /v1/models, filtered for the authenticated API key. A model visible in the Hub is not necessarily callable by every key.
Was this page helpful?