The partnership will see Alibaba deploy its Qwen large language model and Alibaba Cloud infrastructure to transform the fan experience through several key technologies :
This deal extends a playbook Alibaba has already executed at the highest levels of global sport. UEFA joins the International Olympic Committee (IOC) and NBA China among the major organizations selecting Alibaba as a core AI partner . For the Milano Cortina 2026 Winter Games, Alibaba Cloud is already deploying Qwen-powered AI assistants, intelligent pin trading, and serving as the sole cloud provider for Olympic Broadcasting Services . By replicating this model with European football, Alibaba gains a six-year showcase for its enterprise AI capabilities in front of a massive global audience.
The second catalyst hit on June 2, when Alibaba’s Qwen research lab officially released Qwen3.7-Plus, a multimodal agent model that unifies vision and language into a single intelligent system . Unlike its text-only sibling, Qwen3.7-Max, this model can process images, video, and screen captures—and then act on that information .
Benchmark results are what caught the market’s attention. Qwen3.7-Plus achieved a Terminal-Bench 2.0 score of 70.3, edging out Claude Opus 4.6 and competing closely with GPT-5.4 on agent-specific tasks . It also scored 79.0 on ScreenSpot Pro, placing it at the front of the open-API GUI agent field for interpreting and navigating graphical user interfaces . The model can autonomously operate apps, fill forms, write code from screenshots, and automate browsers—capabilities that translate directly into real-world enterprise productivity .
Available immediately through Alibaba Cloud’s Bailian platform (accessible internationally as Model Studio), the model is priced at $0.4-$1.2 per 1M input tokens and $1.6-$4.8 per 1M output tokens . This pricing and availability signal an intent to commercialize the technology quickly, rather than keeping it confined to research benchmarks.
Taken in isolation, either announcement might have generated a modest bump. Together, they reshaped the narrative around Alibaba’s heavy AI spending. Investors have long tracked Alibaba’s massive capital expenditure on AI infrastructure and data centers, treating it as a near-term cost. The UEFA deal converts that spending into a multi-year, revenue-generating contract with a high-profile global partner. The Qwen3.7-Plus launch demonstrates that Alibaba’s homegrown AI can now compete at the frontier with Western counterparts like Anthropic and OpenAI .
The shift is from a cost story to a revenue story . By locking in marquee sports partnerships and proving its AI is ready for prime time, Alibaba is giving investors a tangible path from infrastructure buildout to commercial returns. The 6.6% Hong Kong rally reflects confidence that the pieces are indeed fitting together.
The partnership will see Alibaba deploy its Qwen large language model and Alibaba Cloud infrastructure to transform the fan experience through several key technologies :
This deal extends a playbook Alibaba has already executed at the highest levels of global sport. UEFA joins the International Olympic Committee (IOC) and NBA China among the major organizations selecting Alibaba as a core AI partner . For the Milano Cortina 2026 Winter Games, Alibaba Cloud is already deploying Qwen-powered AI assistants, intelligent pin trading, and serving as the sole cloud provider for Olympic Broadcasting Services . By replicating this model with European football, Alibaba gains a six-year showcase for its enterprise AI capabilities in front of a massive global audience.
The second catalyst hit on June 2, when Alibaba’s Qwen research lab officially released Qwen3.7-Plus, a multimodal agent model that unifies vision and language into a single intelligent system . Unlike its text-only sibling, Qwen3.7-Max, this model can process images, video, and screen captures—and then act on that information .
Benchmark results are what caught the market’s attention. Qwen3.7-Plus achieved a Terminal-Bench 2.0 score of 70.3, edging out Claude Opus 4.6 and competing closely with GPT-5.4 on agent-specific tasks . It also scored 79.0 on ScreenSpot Pro, placing it at the front of the open-API GUI agent field for interpreting and navigating graphical user interfaces . The model can autonomously operate apps, fill forms, write code from screenshots, and automate browsers—capabilities that translate directly into real-world enterprise productivity .
Available immediately through Alibaba Cloud’s Bailian platform (accessible internationally as Model Studio), the model is priced at $0.4-$1.2 per 1M input tokens and $1.6-$4.8 per 1M output tokens . This pricing and availability signal an intent to commercialize the technology quickly, rather than keeping it confined to research benchmarks.
Taken in isolation, either announcement might have generated a modest bump. Together, they reshaped the narrative around Alibaba’s heavy AI spending. Investors have long tracked Alibaba’s massive capital expenditure on AI infrastructure and data centers, treating it as a near-term cost. The UEFA deal converts that spending into a multi-year, revenue-generating contract with a high-profile global partner. The Qwen3.7-Plus launch demonstrates that Alibaba’s homegrown AI can now compete at the frontier with Western counterparts like Anthropic and OpenAI .
The shift is from a cost story to a revenue story . By locking in marquee sports partnerships and proving its AI is ready for prime time, Alibaba is giving investors a tangible path from infrastructure buildout to commercial returns. The 6.6% Hong Kong rally reflects confidence that the pieces are indeed fitting together.