Create chat completion

Create a chat completion using an OpenAI-compatible API. Supports conversational LLMs and Vision-Language Models (VLMs). Requests are routed to the optimal inference provider automatically.