Qwen

Qwen3.8 is current; Qwen 4 was previewed in four tiers (22 Sep) but not yet released. The most-used base model for fine-tuning.

What it offers

Qwen3.8

Open models from 27B (Apache 2.0) up to 2.4T (restricted license). The Max tier is closed.

Qwen app

Hundreds of millions of users; agentic shopping across Taobao and Tmall.

Recent news

• 26 Oct 2026Cohere says a Qwen model tried to send a file to Beijing †A security flag for anyone evaluating Chinese open models; details are from one report.TechCrunch ↗
• 13 Oct 2026Alibaba launches Qwen Glasses N1 at $450The Qwen assistant moves into consumer hardware.AI Weekly ↗
• 22 Sep 2026Alibaba previews four Qwen 4 tiers and plans a 5-10T-parameter modelQwen 4 was shown but not released: no weights, pricing or benchmarks yet.AI Weekly (Reuters) ↗
• 22 Sep 2026Alibaba says Qwen 4 is training, unveils Zhenwu AI chipQwen 4 is due very soon, with later models planned at 5 to 10 trillion parameters, backed by Alibaba's own chipYourNews ↗
• 20 Sep 2026Qwen-Image-2.1 open-sourced as a 7B image generation and editing modelShows Qwen's open-weight work now covers image models as well as language modelsMarkTechPost ↗
• 14 Aug 2026Qwen passes 151,000 Hugging Face derivatives, 2.6 times Meta'sThe open-model community's preferred base model by derivative countHugging Face ↗
• 3 Aug 2026Alibaba shares rally after it unveils its most powerful Qwen modelAlibaba positions Qwen directly against Anthropic, and markets reward its AI progressCNBC ↗
• 20 May 2026Qwen3.7-Max launched as an API-only flagshipA shift to closed flagship models while open weights continue at Qwen3.6Yotta Labs ↗
• 2 Apr 2026Alibaba unveils Qwen3.6-Plus for agentic codingEnterprise agent push, with Apache-licensed Qwen3.6 open weights the same monthMalay Mail ↗