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
- [Bench] Fix bench_serving multi-turn reasoning handling and add --multi-turn-echo-reasoning
- [HiCache] Add fast_file local storage backend (direct host-buffer I/O, parallel reads, background eviction)
- [Anthropic] Fix /v1/messages usage cache field mapping
- [Bug] /v1/messages never reports cache_creation_input_tokens, so cache writes are billed as plain input
- [Bug] FlashKDA prefill returns a tuple and crashes GLM-5.3-Flash with radix cache disabled
- [Spec] Reduce DSpark CUDA-graph memory probe over the capture group
- Add a CUDA event pool for hot-path wait_stream calls
- [Benchmark] Record radix cache protocol in bench_serving JSONL rows.
- [Bug] EAGLE draft-extend plan stream lacks an input-producer dependency
- [Bug] A single /v1/chat/completions request with top_logprobs > vocab size kills the scheduler
- Docs
- Python not yet supported