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
- [Feat] deepseek-v4.1: Encoder-only prefill for PD (--dsv41-encoder-only-prefill)
- [Feat] [NVIDIA] megakernels: Qwen3.5 binding and EAGLE entry wiring for NVIDIA megakernel plugins
- [Doc] [AMD] cookbook: Align DeepSeek-V4 MI355X AgentX recipe with InferenceX #4315
- [Fix] qwen3_coder: don't constrain non-strict tools under tool_choice="auto"
- [sglang-miles] [Feat] beam: Port Beam BF16 serving and LoRA
- [Perf] [NVIDIA] DSA: Pick any divisor of the tile count for the TileLang sparse decode split
- ci: run only the A5 accuracy job in nightly-test-npu
- [Fix] qwen3_coder: match <tool_call> as a token via xgrammar token_markers
- [Feat] eplb: Non-blocking expert rebalance
- [Chore] deepseek-v4.1: Harden the prefill graph capture of the full top-k break
- Docs
- Python not yet supported