Alibaba is reportedly considering a revenue-sharing model for major users of its upcoming Qwen3.8-Max open-source AI model, according to sources. This approach mirrors a strategy previously used by Moonshot, which required partners offering its Kimi K3 model as a service to share a portion of their revenue. The Qwen3.8-Max model, like Moonshot's Kimi K3, is described as an open-source, open-weight model, allowing developers to run or adapt the system. However, the new model appears to include terms that could require users to pay for collaboration and early access to future revisions.

The shift in monetization strategy for open-source AI models comes as companies increasingly rely on such models for commercial applications. Alibaba's Qwen3.8-Max is positioned as a major release, following the success of Moonshot's Kimi K3. The licensing terms for Moonshot's Kimi K3 required partners generating more than $20 million annually to share up to 30% of their revenue with Moonshot. This model has now been adapted by Alibaba, with sources indicating that similar provisions may be included in the Qwen3.8-Max licensing agreement.

The Qwen3.8-Max model is described as a 3.8 version of Alibaba's Qwen series, which has been gaining traction in the AI industry. The model's open-weight nature allows developers to fine-tune and optimize the system for specific applications. However, the new revenue-sharing model introduces a financial obligation for major users, which could affect how companies deploy and scale the model. This approach may also influence how other open-source AI model providers structure their licensing agreements in the future.

The revenue-sharing model could lead to increased costs for companies using the Qwen3.8-Max model, potentially affecting their deployment strategies. It may also create vendor lock-in, as users may be reluctant to switch to alternative models if they are required to pay a share of their revenue. Additionally, the model's governance structure could become more complex, with users needing to navigate licensing terms and revenue-sharing agreements. Market reactions could vary, with some companies embracing the model for its capabilities while others may seek alternatives to avoid financial obligations.

The approach taken by Alibaba with the Qwen3.8-Max model highlights the evolving landscape of open-source AI monetization. While open-source models have traditionally been associated with low or no cost, the inclusion of revenue-sharing terms suggests a shift toward more commercialized models. This could influence how other companies in the AI industry structure their licensing agreements, potentially leading to a new standard for monetizing open-source AI models. The long-term impact of this strategy remains to be seen, but it signals a growing trend of integrating financial incentives into open-source AI initiatives.