Ant Group’s Lingbo plans record financing deals to cash in on robotics rush

Ant Group’s (HK: 6688) Lingbo Technologies recently started its first external funding round, aiming to raise 1.5 billion yuan ($206 million). 

The company intends to hold a second funding round before the end of 2026, which would make it the fastest funding round of any startup in China’s robot-brain sector. 

Why is Lingbo focusing on robot brains? 

Lingbo Technologies, which has been fully funded by its parent company Ant Group since its launch, has started its first outside fundraising round and is looking to raise 1.5 billion yuan (about $206 million). The company has shared its plans to hold a second funding round before the end of 2026. 

The decision to seek outside money is due to the budget pressure that Ant Group is currently facing on computing resources. The company is also pouring money into AI projects in other parts of the business. Notably, analysts say Ant has already put more money into Lingbo than a typical top startup would raise in its early rounds. 

Lingbo’s main focus is on building the “brain” for robots. This means the software and AI models that help robots perceive the world, make decisions, and take actions. Lingbo is betting that a single, powerful “brain” can work with many different robot bodies. 

The company calls this strategy “one brain, many machines,” and it is similar to the strategy of a U.S. company valued at over $11 billion despite not making many robots of its own, called Physical Intelligence. 

The company repurposes technology built for Alipay, Ant’s mobile payment app, which includes systems for high-speed transactions and making decisions in real-time, even with uncertainty, to help robots read and act in the physical world.

Lingbo’s model stack, called LingBot, covers many areas, including vision, depth perception, mapping, and video understanding. Its flagship model, LingBot-VLA 2.0, was trained on about 60,000 hours of real-world data from robots doing physical tasks. The model now works with more than 20 different robot configurations from 17 brands, including Unitree and AgiBot. 

Lingbo’s models have also gained significant attention on GitHub, with nearly 30,000 stars, about 2.7 times more than the second-ranked domestic team. 

Lingbo has so far received strong interest from investors. Some investment firms reportedly reached out even before the funding round officially began. At the World Artificial Intelligence Conference in Shanghai, some investors went straight to Lingbo’s booth to check out the technology and show interest.

Roughly half of the capital going into embodied intelligence now flows to companies building the “brain,” the perception and decision models, rather than the physical hardware, and that share is climbing. 

The Chinese government has made robotics and “embodied intelligence” a top priority in its latest 15th Five-Year Plan, which provides billions of dollars in subsidies and support for the industry. 

Cryptopolitan reported that Ant Group itself has led funding rounds for at least 12 robotics companies since early 2025, including a $73.58 million round for humanoid startup Zeroth. 

Can Lingbo justify its high valuation? 

Despite the company’s fast growth and the mounting investor interest, Lingbo’s Chief Scientist, Shen Yujun, has been honest about the challenges ahead. He said robotics has not yet had its “GPT-1 moment” due to a lack of the millions of hours of data needed for a true breakthrough.

After the fundraising, Lingbo will remain Ant’s main business in embodied and physical AI. The parent company will continue to provide support on infrastructure, data, and deployment. 

Insiders say the firm’s final valuation will depend on several factors, including whether or not LingBot models show a clear ability to work well across many different types of robots and tasks. 

The company must show that it is ready for large-scale deployment, and it must also create a working “data flywheel,” where more usage leads to more data, which makes the models better, leading to even more usage. Lastly, the company would need to grow from an internal Ant Group project into a true global platform for physical AI.