LMSYS releases SGLang + RadixAttention

SGLang + RadixAttention released

Why it matters

Brought a cache layer for workloads with overlapping prompts, such as multi-turn and agents, down to the runtime level

Source — Primary sources first; secondary reports are labelled “Press”

Officiallmsys.orghttps://www.lmsys.org/blog/2024-01-17-sglang/

Added in Week 35, 2026 · See that week

Same week

DistServePaperAlphaGeometry paper published in NaturePaperPinecone serverless announcedOfficial

Nearby on the same axis

DistServePaperSGLang paperPaperSarathi-ServePaperSplitwisePaper