You.com launches Inference Router, an OpenAI-compatible API for open-weight models with built-in web search

You.com CEO Richard Socher announced Inference Router, a hosted inference service with one OpenAI-compatible API and no platform fee. Existing OpenAI-compatible code switches by changing the base URL.

You.com says the service offers:

• a minimum 99.9% uptime SLA backed by service credits;
• zero data retention across every model call and web search request;
• 100% US-hosted infrastructure with multi-provider failover;
• built-in web search that runs on the same server as the model, which You.com says cuts latency by about 15% in its own testing.

Models come from DeepSeek, Kimi, Z.ai, Qwen, MiniMax, Meta and others, including Kimi K3, GLM 5.3 and Qwen3.8 Flash. You.com cites an independent Braintrust evaluation in which an open-weight model paired with its web search matched or beat a leading closed model using its own built-in search, at up to 75% lower cost per question. That is You.com's claim.