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
- [Fix] GLM: Preserve NextN PTPC quantization for excluded experts
- [Fix] [AMD] quark: Skip no-copy loads after MXFP4 quantization
- [Fix] parser: Fix reasoning/tool-call grammar handoff and GLM-4.5 argument parsing
- [Fix] dsv4: Mirror SGLANG_DEFAULT_THINKING between prompt and reasoning parse sides
- [Feat] ultravox: Add Ultravox speech LLM support
- [Perf] [NVIDIA] deepseek-v4.1: Fuse BF16 prefill indexer score epilogue
- [Refactor] router: align Generate forwarding with the rust server API
- [Fix] moe: scale the KT routed output once when the top-k already folds the factor
- [Perf] deepseek-v4: Reuse synchronized prefill tail counts
- [Fix] scheduler: Keep the priority preemption order consistent across TP ranks
- Docs
- Python not yet supported