Humanity Unfolds

llm-d 1.0 brings open inference to production operations

Oct 20, 2026

The llm-d project releases version 1.0 with a stable v1 configuration API, multi-tenant scheduling, coordinated upgrades and session-aware serving for agent workloads.

Open infrastructure connects model servers such as vLLM and SGLang with distributed routing across clouds and accelerators. The Apache-2.0 project is a community effort, not a model release or a single vendor product.

Explore on Humanity Unfolds

People

Sources