Qwen3.8 models released with open Apache 2.0 weights
Alibaba’s Qwen AI team has released open weights for the Qwen3.8 model family under the Apache 2.0 license. The main model, Qwen3.8-27B, is a dense multimodal system with 27 billion parameters. Qwen says it surpasses the larger Qwen3.7-Plus in coding and office-related tasks, while also offering stronger agent capabilities for independent planning and more reliable task completion.
Qwen3.8-27B natively supports context windows of up to 262,000 tokens. With the YaRN method, the context can be extended to one million tokens. The model handles text, images and video, including diagrams, documents and multi-hour recordings. Its flexible thinking mode is enabled by default but can be switched for individual queries.
The team has also published weights for Qwen3.8-2.4T-A95B, a much larger model intended to run at the Max level. Both models are available through Hugging Face and ModelScope. A hosted version supporting one million tokens of context is expected to arrive soon on Qwen Cloud, Alibaba’s AI service.
Written by Sakhanda Wire AI from reporting by The Decoder. Facts, quotes and full context are in the original article below.