We implement an OpenAI Chat Completions compatible endpoint. To get started quickly with chat inference, check out the /v1/chat/completions endpoint.