LMSYS launches SGLang Alliance with Nvidia, AMD and others

On 2026-11-12 at 08:00, LMSYS Org announced the formation of the SGLang alliance. NVIDIA, AMD, Intel, Google, Meta, Cloudflare, Modal, and RadixArk have joined as hardware, model, and infrastructure partners, working alongside the SGLang maintainers to advance open source AI inference. The alliance states that its goal is to make inference more efficient in order to serve a growing AI ecosystem. At this stage, the publicly available information stops at the alliance's announcement: the material does not detail any follow-up collaboration arrangements or implementation plans among the participating parties, so no further progress or concrete next steps have been reported yet. The announcement frames the initiative as a joint effort, with the participating organizations contributing across hardware, model development, and infrastructure, while the SGLang maintainers continue their work on the open source inference stack. No timeline, benchmark targets, or milestone deliverables are provided in the available reporting.