Alibaba Qwen Monetization: How the $53B AI Bet Turns Free Models Into Revenue

Video generation credits are billed separately, with prices ranging from 26 yuan to 968 yuan depending on the quota package . Basic access to the Qwen chatbot remains free, but the office assistant operates on a usage quota. Once exhausted, users must pay to continue . The paid plans also grant access to the flagship Qwen3.8-Max model and premium features like scheduled tasks and the thinking/research module .

Separately, a different team within Alibaba launched QwenWork in public beta around the same time, starting at 78 yuan/month .

2. Qwen3.8-Max Revenue-Sharing for Large Commercial Users

On August 7, 2026, Reuters reported that Alibaba plans to require major commercial users of its upcoming Qwen3.8-Max open-weight model to share a portion of revenue . The model is released as open-source/open-weight — developers can download the core settings and run them on their own servers — but large-scale commercial deployments will trigger a commercial agreement .

Moonshot's precedent: Moonshot's Kimi K3 license already requires any partner generating over $20 million in annual sales from the model to negotiate a commercial agreement, with a revenue share of up to 30% . Alibaba's exact terms are still under negotiation, but the structure mirrors this approach — targeting high-revenue users rather than small developers .

The Qwen3.8-Max preview (with 2.4 trillion parameters) was first released on July 19 alongside Alibaba's Token Plan individual subscriptions .

3. The $53 Billion Infrastructure Bet

In February 2025, Alibaba announced plans to invest at least 380 billion yuan (~$53 billion) over three years in AI and cloud infrastructure — more than it had spent on AI and cloud in the entire previous decade . By May 2026, CEO Eddie Wu confirmed the company would exceed that target due to early signs of returns . Spending is split roughly evenly across three years .

By September 2025, Alibaba had already signaled it would add to the original plan, aiming to launch its first data centers in Brazil, France, and the Netherlands .

4. Financial Pressure and Price Wars

Chinese AI companies are under intense financial strain from the capital-intensive race to train larger models, even as they slash prices to capture market share. Alibaba has run aggressive promotions — including a 3-billion-yuan Lunar New Year promotional budget and deep discounts on its Token Plan, with coding agents as low as $6/month — while simultaneously moving flagship models like Qwen3.6-Max and Qwen3.7-Plus from open-source to closed, API-only paid access .

The industry-wide dynamic: burn cash to win users, then scramble to monetize. Alibaba's three-front strategy for Qwen is the clearest signal yet that Chinese AI labs are settling on a commercial model — even if it means putting a paywall on what was once free.