LMSYS launches SGLang Consortium with NVIDIA and AMD

LMSYS Org launched the SGLang Consortium alongside founding members including NVIDIA, AMD, Intel, Google, Meta, Cloudflare, Modal, and RadixArk to guide the roadmap of the open-source AI inference engine. Announced at the inaugural SGLang Summit in San Francisco, the forum will facilitate joint engineering work on frontier model support, multimodal workloads, and hardware optimization. SGLang maintainers will retain responsibility for technical decisions and code acceptance. Open article →