Chinese technology giant Alibaba is preparing to introduce a revenue-sharing model for major commercial users of the next version of its Qwen open-source AI model, signalling a significant change in how leading AI companies plan to monetize powerful open-weight models. According to sources familiar with the matter, the upcoming Qwen3.8-Max model will remain openly available, but enterprises generating substantial revenue from it will be required to enter commercial licensing agreements with Alibaba.
The move reflects a broader shift in the artificial intelligence industry, where companies are increasingly adopting “freemium” business models—offering models freely to developers and researchers while charging large-scale commercial users.
What Will Change?
Under the proposed model:
- The AI model’s weights will remain openly available.
- Developers and smaller organizations will continue to access and deploy the model without charge.
- Large enterprises that generate significant commercial revenue using the model will be required to share a portion of that revenue with Alibaba through commercial agreements.
Alibaba has not publicly disclosed the revenue threshold or the percentage that large users may be required to pay.
Inspired by China’s Emerging AI Business Model
Reuters reports that Alibaba’s strategy resembles the licensing model recently introduced by Chinese AI startup Moonshot AI for its Kimi K3 model.
Under Moonshot’s approach, companies generating more than $20 million in annual revenue from the model must enter revenue-sharing agreements, with some arrangements reportedly reaching 30% of revenue. Alibaba’s exact commercial terms, however, are still under discussion.
From Free Deployment to Enterprise Monetization
Previously, Alibaba allowed organizations to deploy Qwen models free of charge on their own infrastructure, charging customers only when they used the models through Alibaba Cloud.
The new approach would expand monetization beyond cloud hosting by requiring large commercial users to contribute revenue even when running the models on their own private infrastructure.
Competition in the Global AI Race
Alibaba’s latest strategy comes as competition intensifies between Chinese and US AI companies.
Chinese firms such as Alibaba and DeepSeek have rapidly improved the performance of their AI models while keeping many of them open-weight, helping attract developers worldwide. At the same time, US companies including OpenAI, Anthropic, and Google continue investing heavily in proprietary frontier models.
Industry analysts say the shift toward revenue sharing could help AI developers recover the enormous costs of training and operating advanced large language models while preserving the benefits of open access for most users.
Balancing Open Access and Sustainability
The proposed licensing model reflects a growing trend in the AI industry to balance openness with commercial sustainability.
Rather than abandoning open-weight models, companies are increasingly allowing developers to innovate freely while charging enterprises that generate significant business value from the technology. This approach enables AI developers to continue funding research, infrastructure, and future model development without restricting broader innovation.
No Official Announcement Yet
Alibaba has not officially announced the new pricing structure, and the plans remain based on information from sources familiar with the matter. The company is expected to provide further details when it unveils the next generation of its Qwen AI model.
What It Means
If implemented, Alibaba’s move could influence how other AI companies commercialize open-weight models. As demand for enterprise AI solutions grows, hybrid licensing models that combine open access with revenue sharing may become increasingly common, allowing developers to innovate while ensuring AI companies have sustainable business models to support future advancements.
