en.Wedoany.com Reported - Alibaba, through its Tongyi Lab, has released the Qwen Robot Suite, a set of models focused on "embodied intelligence," with the core goal of enabling machines to perceive space, reason, and take action.

This move follows Nvidia's release of the physical AI model Cosmos 3, further validating Nvidia CEO Jensen Huang's assessment that China's developer ecosystem is less affected by chip restrictions, while the Western focus remains on data center power demands.

The Qwen-Robot Suite comprises three core models: Qwen-RobotManip (a generalizable vision-language-action model), Qwen-RobotNav (a scalable vision-language navigation model), and Qwen-RobotWorld (a video world model designed for embodied intelligence). Alibaba claims that the model, which uses the lighter Qwen3.5-4B (rather than the trillion-parameter Qwen 3.7 Max), scored 59.83 on the RoboChallenge real-world robot benchmark, with a task success rate of 45%.

R&D in China's AI robotics sector is accelerating. Companies such as Tencent, Unitree, AgiBot, UBTech, Galbot, Spirit AI, GigaAI, as well as electric vehicle makers Xpeng and Xiaomi, are all involved. Alibaba's wholly-owned subsidiary, the South China Morning Post, noted that embodied intelligence is rapidly becoming the next frontier of global AI.

Nvidia is attempting to play the role of an "enabler" in this field rather than a direct competitor. Its open-source models aim to replicate the success path of CUDA in the graphics processor market, using products like Cosmos, GR00T, and Isaac to ensure future robotics platforms are built around Nvidia's hardware and software stack. Alibaba's release can be seen as an intention to build a similar ecosystem for Chinese robotics companies, especially against the backdrop of the Chinese government's informal emphasis on reducing reliance on American hardware or software.

Since Nvidia does not operate in China, direct comparisons between the two products are difficult, and the scale difference is significant: Cosmos 3 is an open-world foundation model with scores reported by multiple vendors but not covering RoboChallenge, while Alibaba's results are self-reported based on a single benchmark. For Nvidia, a fact repeatedly emphasized by its CEO is that China, feeling aggrieved by U.S. AI policies, tends to build its own ecosystem from scratch rather than integrating chips, models, or even open-source solutions. This trend could lead to Nvidia losing its second-largest profit market, with the robotics segment alone potentially costing the company billions of dollars in revenue.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com