Alibaba open-sources Qwen3.8-27B: 27B multimodal model beats Qwen3.7-Plus, Apache 2.0
Alibaba_Qwen · x · 2026-08-14
Alibaba's Qwen team announced open weights for Qwen3.8 series: Qwen3.8-27B is a native multimodal dense model that outperforms Qwen3.7-Plus overall with just 27B parameters, excelling in real-world coding and office workflows. It supports 262K native context, extendable to 1M via YaRN, and is Apache 2.0 licensed. The Max-level Qwen3.8-2.4T-A95B open weights were also released. Developers can deploy 27B locally or build agents with 2.4T.
Related event: Alibaba Open-Sources Qwen3.8-27B, Outperforming Plus(2 posts)→
More from Models
- Qwen3.8-Max open weights now on Nebius Token Factory Day 0 — Arindam_1729 · 2026-08-14
- Qwen 3.8-Max Now on Modal: 2.4T Params, 1M Context, Custom DFlash Speculator — AAAzzam · 2026-08-14
- GLM-5.3 Released: Post-Trained on 743B Base, Big Leaps in Coding and Cyber Defense — sudoraohacker · 2026-08-14
- Qwen3.8 27B vs Qwen3.6 27B vs Gemma 4 31B: Comparing 24GB GPU Options — MaySaki2 · 2026-08-14
- Qwen3.8-27B open-sourced with SGLang Day-0 support, 206 tok/s on RTX 5090 — ying11231 · 2026-08-14
- Qwen3.8-Max launches on DigitalOcean Serverless with 1M context — Alibaba_Qwen · 2026-08-14