sglang
https://github.com/sgl-project/sglang
Python
SGLang is a high-performance serving framework for large language models and multimodal models.
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported1 Subscribers
Add a CodeTriage badge to sglang
Help out
- Issues
- [HiCache] write_back eviction blocks the scheduler for hundreds of ms per call under steady load, lowering the prefix-cache hit rate compared with write_through
- [Feat] weight_cache: Add the vmm_fd transport
- [Fix] scheduler: backport no_grad CUDA graph capture (#43854) to amd_march
- [Doc] [XPU] qwen3: Add the intel_xpu attention backend and XPU graphs to the BMG recipe
- [Perf] [AMD] aiter: Speed up asm fp8 MLA prefill for 12-head and mixed-prefix batches
- [Perf] [AMD] kda: Run gfx950 KDA prefill extend on AITER Gluon chunk_kda
- [Feat] hicache: Add --hicache-mamba-size to pin the HiCache Mamba host pool
- 【WIP】NPU Graph refactor
- [Fix] auth: Warn when ordinary --api-key also unlocks ADMIN_OPTIONAL endpoints
- [Feat] cohere-asr: Support CohereLabs/cohere-transcribe-03-2026
- Docs
- Python not yet supported