SGLang reaches v1.0 with a consortium backed by NVIDIA, AMD, Intel, Google and Meta
TL;DR The first SGLang Summit brought v1.0, Miles v0.2 and a consortium with nine founding members, including competing chipmakers.
In plain words: an inference engine is the server software that actually runs an AI model and answers requests. SGLang is one of the two most used open-source engines (with vLLM). Rival chip companies now back it together.
What happened
SGLang v1.0, Miles v0.2 and the SGLang Consortium were announced, with nine founding members: NVIDIA, AMD, Intel, Google, Meta, Thinking Machines Lab, Cloudflare, Modal and RadixArk.
Who is affected
Anyone serving open models in production or choosing an engine for a new deployment.
Why it matters
It is rare for competing chipmakers to back the same inference engine on the same day. It signals SGLang as a common base for serving open models in production, next to vLLM.
What to do
Nothing urgent. If you are on an older SGLang, read the v1.0 notes before upgrading.