Alibaba Plans Revenue Sharing for Major Commercial Users of Its Open-Source Qwen AI Model

Alibaba Introduces Revenue Sharing for Qwen AI Model Commercial Users

Alibaba plans to charge big users of its open-source AI model, specifically targeting commercial deployers of its upcoming Qwen3.8-Max release who generate significant revenue1

1

2

2

. The Chinese tech giant will implement a revenue-sharing arrangement for major users of the next version of its Qwen AI model, according to two people familiar with the company's strategy2

2

. While the Qwen3.8-Max model remains open-weight, meaning its underlying learned settings stay available for download, the largest commercial users will now need to negotiate commercial agreements with Alibaba3

3

4

4

. This shift breaks from Alibaba's previous approach of letting companies run its open-source models on their own infrastructure without payment5

5

.

Chinese AI Firms Converge on Freemium Business Model

The move follows a licensing policy pioneered by Chinese AI startup Moonshot with its Kimi K3 model, which requires anyone offering the model for sale as a service and generating more than $20 million in annual sales to establish a commercial agreement2

2

4

4

. Moonshot's arrangement can demand up to 30% revenue sharing from partners exceeding this threshold, according to sources familiar with the terms2

2

. Chinasoft International, a Chinese IT services provider, disclosed a revenue-sharing agreement with Moonshot in a regulatory filing last month, though it did not specify the percentage5

5

. Alibaba also plans to ask for a revenue share, though the specific rate remains under negotiation as discussions continue2

2

. Paddy Srinivasan, CEO of cloud computing firm DigitalOcean Holdings, confirmed his company has a commercial agreement with Moonshot and described this approach as "a tried and tested open-source freemium model"2

2

5

5

.

Qwen3.8-Max Performance and Market Position

Alibaba released Qwen3.8-Max earlier this week, describing it as its most capable model with 2.4 trillion total parameters and 95 billion active parameters4

4

. Early benchmarking suggests the model excels in areas such as agentic coding, where bots write and fix code unassisted1

1

. Alibaba has claimed Qwen ranks as the world's number-two open-weight design, giving it substantial reach that makes monetizing open-source models plausible3

3

. The company's stock responded positively to the announcement, rising 4.5% in premarket trading in New York and 7% on the Hong Kong exchange4

4

. Alibaba chief Eddie Wu believes AI models and services will constitute half of external cloud income by 2027, representing approximately a tenth of total sales1

1

.

Economic Pressures Drive Monetization Strategy

The push to monetize open-source models comes as Chinese AI firms face intense economic pressures from multiple directions3

3

. A brutal price war has raged among China's AI companies, with DeepSeek making a 75% discount permanent and squeezing competitors' ability to charge for access3

3

. Meanwhile, compute costs remain enormous—Moonshot reportedly used 20,000 Nvidia chips from Alibaba's cloud computing infrastructure for training3

3

. Alibaba's capital expenditure of $18.3 billion in its last accounting year left the company with negative free cash flow, though inflows are expected to return this year1

1

. The Kimi K3 model costs about a third of Anthropic's Fable model according to listed prices of input and output tokens, illustrating the pricing pressure Chinese AI firms face against US competitors2

2

.

Model as a Service and Cloud Computing Advantages

Alibaba's fast-growing cloud computing division provides opportunities to monetize its models through a "model as a service" approach, similar to Microsoft's strategy1

1

. The company offers smaller companies cloud-based access to pre-trained models, which they can deploy as chatbots or other applications, paying according to usage1

1

. To date, Alibaba has charged developers for using its models when hosted on its own cloud platform but allowed most open-source offerings to be used in customers' own data centers without payment2

2

. Dan Fu, vice president of kernels at Together AI, explained that AI software providers generate revenue by optimizing service efficiency around tokens, the fundamental units AI systems use to process queries: "At the application layer, there's value out there for how you use it, how you actually get the models and the tokens to do something useful"2

2

5

5

.

Geopolitical Considerations and US Chip Restrictions

The bifurcated tech world that splits China from the US presents both challenges and opportunities for Chinese AI firms1

1

. US chip restrictions limit China's access to advanced chips and other technology, forcing innovation—like Google, Alibaba has developed its own proprietary chips1

1

. US hardware makers need local partners to operate in China, with Apple winning approval to launch AI-embedded devices using local technology including Qwen1

1

. China's labs have leaned on openness partly to spread their models worldwide while US leaders at OpenAI and Anthropic kept their best systems closed3

3

. However, charging heavy commercial users risks blunting this advantage3

3

. Deals are taking shape under which US firms share revenue generated by Chinese models with the Chinese AI labs that created them, even as the White House has accused Moonshot of stealing technology from Anthropic—a claim Chinese officials have rejected as unfounded2

2

.

Broader Open-Source AI Landscape

The open-source AI movement continues expanding, with US firms joining the competition5

5

. Thinking Machines Lab, a San Francisco-based AI startup founded last year by OpenAI's former Chief Technology Officer Mira Murati, released its first open-source model last month and is widely expected to follow with more powerful versions2

2

. Lin Qiao, CEO and co-founder of Silicon Valley-based Fireworks AI, stated: "I don't see a fundamental barrier" to powerful open-source U.S. models, adding "We are really waiting for that to happen"5

5

. For Alibaba, the revenue-sharing strategy represents a bet that reach can eventually convert into sustainable income, having spent heavily to make Qwen popular, the company now aims to collect from businesses that have built commercial products on top of it3

3

.