Qwen ships Qwen3.8-27B under Apache 2.0 — a 27.8B dense multimodal model claiming 61.7% on SWE-Bench Pro
AI Release Tracker·high signal
Alibaba's Tongyi Lab released Qwen3.8-27B on 2026-08-14: 27.78B dense parameters, text/image/video input, Apache 2.0, and a native 262,144-token context extensible to 1M via YaRN. Reported scores include SWE-Bench Pro 61.7%, LiveCodeBench 90.3%, GPQA Diamond 89.2%, and Terminal-Bench 2.1 at 73.0 (up from 63.4 in Qwen3.6-27B), with DeepSWE 1.1 jumping 13.3 → 42.2. It arrived 11 days after Qwen3.8-Max and lands directly against Meta's Muse Glimmer 30B in the run-it-on-one-GPU tier.