Qwen3.8-27B Open Source: 206 tok/s on RTX 5090 with SGLang Day-0 Support
cedric_chee · x · 2026-08-15
Alibaba's Qwen3.8-27B is now open source, with Day-0 support from SGLang. On a single RTX 5090, it achieves 206.1 tok/s decode speed using NVFP4 and DSpark, and 38.28 tok/s on DGX Spark. The model excels in agentic planning and long-horizon tasks, earning the title 'king of small models'.
Related event: Alibaba Open-Sources Native Multimodal Qwen3.8-27B Model(48 posts)→
More from Models
- 200 chess games suggest in-context learning doesn't work yet, experimenter says — akbirthko · 2026-10-16
- OpenAI's internal model shows real math "research taste," experts say its proofs go beyond prior literature — cong_ml · 2026-10-16
- User catches Opus spinning up throwaway zz- debug scripts, echoing the METR OAI-HuggingFace incident — holdenmatt · 2026-10-16
- OpenAI drops 722 math papers tied to unreleased model, mathematicians demand receipts — thursdai_pod · 2026-10-16
- Fable 5/5.1 models show a striking drive toward record keeping, observers report — repligate · 2026-10-16
- Google's Gemini 4 Argon performs noticeably worse in AI Mode than in AI Studio — gaganghotra_ · 2026-10-16