What happened
DeepSeek is set to lead a seed funding round for Winverse, a robotics AI startup founded in 2026 by Hu Han, a former lead scientist at Tencent. According to a report from Investing.com citing Bloomberg, the round values the startup at approximately $400 million. Other investors, including Meituan’s Long-Z Investments, have also committed to the financing. The deal remains unfinalized, and specific terms regarding valuation and participation are subject to change.
DeepSeek is reportedly leading a seed funding round for Winverse, a startup established in 2026 by Hu Han, who previously led vision-language and embodied teams at Tencent. The round is valued at $400 million, with participation from investors including Meituan’s Long-Z Investments.
The deal is not yet finalized, meaning terms such as the exact valuation and investor participation could shift. This investment represents one of DeepSeek's most significant startup bets since its founding in 2023.
Hu Han, a Tsinghua University PhD graduate with experience at Microsoft Research Asia and Baidu, left Tencent during a period of internal restructuring. His new venture, Winverse, focuses on developing AI specifically for robotics, moving beyond the capabilities of standard large language models.
Source details: ng.investing.com ↗
Why it matters
This investment signals a strategic shift for DeepSeek as it moves beyond large language model development into embodied AI and robotics. By backing Winverse, DeepSeek aims to bridge the gap between its existing AI models and real-world applications, such as humanoid robotics on factory floors. This move highlights the growing competition among Chinese AI firms to secure specialized talent and technology for physical-world automation, a sector where US startups like Skild AI and Physical Intelligence have already achieved multi-billion dollar valuations. The involvement of a former Tencent scientist underscores the trend of top-tier researchers leaving established tech giants to launch specialized labs focused on world models and robotics.
The investment marks a pivot for DeepSeek, which has primarily focused on model training and data centers, toward the physical application of AI. By targeting robotics, the company is attempting to capture value in the 'embodied intelligence' sector, which requires different technical approaches than text-based .
The move reflects a broader trend in the Chinese AI ecosystem where researchers are departing major tech companies to form specialized labs. This is occurring as Chinese firms seek to close the gap with US-based robotics AI leaders like Skild AI and Physical Intelligence, which have secured significant capital and backing from major US tech firms.
DeepSeek’s involvement is part of a larger capital-intensive strategy, as the lab is also reportedly securing at least 80 billion yuan ($12 billion) in its own funding round ahead of a planned 2027 market debut.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
An agent must create a draft calendar event for Tuesday at 2 p.m. Which evidence would establish the requested result?
What to watch next
Observers should monitor the finalization of the funding round and whether DeepSeek’s involvement leads to tangible co-development projects, similar to its previous stake in Unitree Robotics. Additionally, the progress of Winverse under Hu Han’s leadership will be a key indicator of whether specialized Chinese robotics AI labs can successfully compete with well-funded US counterparts. The broader market will also watch for further consolidation or spin-offs of AI talent from major Chinese tech firms as they restructure to address the demand for embodied intelligence.
The primary unknown is the final structure of the deal and whether it will include specific technical collaboration agreements between DeepSeek and Winverse.
Market analysts will be watching to see if this investment leads to a series of similar moves by DeepSeek to build an ecosystem of specialized AI startups, potentially challenging the dominance of established robotics hardware manufacturers.
The success of Winverse will serve as a test case for whether Chinese startups can effectively commercialize embodied AI models in the face of thinner investment liquidity compared to the US market.