Audience
Enterprise engineering and platform teams running AI in production
About Xinference
Xinference is an enterprise AI inference platform for teams that want to run open models without building the serving stack themselves. Most teams start on the Model API: 300+ open models behind one OpenAI-compatible endpoint hosted in Australia. Switching from an existing provider takes two lines of code. As usage grows, the same workloads move to Dedicated Inference on reserved GPUs, or to a private deployment inside the customer's own cloud or data centre. Every deployment includes one control plane: per-request logs, live TTFT and TPOT monitoring, role-based access control, audit logs and SSO. Xinference does not train on customer data and does not retain it by default. Common workloads include enterprise RAG, customer assistants, agents and function calling, coding assistance, document extraction, speech and image generation.
