Alibaba open-sources Qwen3.8-Max and 27B models with revenue share
Alibaba drops open weights for Qwen3.8-Max and 27B starting August 11
Alibaba is set to release the open weights for its flagship Qwen3.8-Max and the more accessible Qwen3.8-27B model during the week of August 11, 2026. This move gives developers full ownership and control over these models, enabling self-hosting, fine-tuning, and deep customization without relying on Alibaba's proprietary API. Qwen3.8-Max is expected to be one of the largest open-weight models available, while the 27B variant targets teams with moderate compute budgets. The release is widely seen as a direct challenge to Meta's Llama 4 and DeepSeek's open-weight ecosystem, particularly in the Asian market where Qwen already has strong adoption.
Alibaba is also introducing a revenue-sharing model for large commercial users of Qwen3.8-Max. While the weights are open, companies generating significant revenue from the model will likely need to share a percentage with Alibaba, creating a hybrid business model that balances open-source accessibility with commercial sustainability. This approach could set a new precedent for how AI companies monetize open-weight releases. For developers, the immediate benefit is unprecedented access to frontier-scale reasoning and coding capabilities. For enterprises, the revenue-sharing clause adds complexity but signals Alibaba's long-term investment in the Qwen ecosystem. Expect rapid community adoption, fine-tuned derivatives, and possible integrations into agentic workflows within weeks of the release.
- Qwen3.8-Max and Qwen3.8-27B open weights expected during week of Aug 11, 2026
- Large commercial users must implement Alibaba's revenue-sharing model for Qwen3.8-Max
- Developers gain self-hosting and customization rights, competing with Llama 4 and DeepSeek
Why It Matters
Open-weight front-runners now include Alibaba, forcing enterprises to factor revenue sharing into their AI stack decisions.