Serving
OpenAI-compatible API
What is openai-compatible api?
An OpenAI-compatible API implements the same request and response shape as OpenAI’s chat completions endpoint. It lets existing client code point at a different model by changing a base URL and a key, with no other modification.
It has become the de facto interface for language model serving, which makes it a portability mechanism as much as a convenience: a stack built against it can move between providers without a rewrite.
Compatibility is a spectrum in practice. Core chat completion is near-universal; streaming, tool calling, structured output and log probabilities vary, and are worth checking against the specific features your client uses.
Related terms
Start with the free tier
A magic link creates your account, your tenant and your first API key. No card until you ask for compute.