Humanoid robot foundation model startup Deta Intelligence has closed nearly 500 million RMB (approximately $70 million) in an Angel++ round, according to 36Kr. The round attracted a mix of listed company strategic investors and financial institutions, though specific names were not disclosed.
The company, which focuses on building the foundational AI models that power humanoid robots, plans to use the capital to continue iterating its core model, accelerate mass production of proprietary data collection hardware, and close the data loop for real-world training. A significant portion will also go toward expanding the core R&D team, with the goal of pushing the technology into actual industrial environments for validation and deployment.
This round — an unusually large Angel++ raise — signals that investors are placing big bets on the foundation model layer of embodied intelligence, even at a very early stage. The fact that multiple listed companies participated as strategic investors suggests that industrial players see Deta's technology as potentially enabling for their own automation roadmaps.
For founders and operators in the embodied AI space, Deta's raise is a signal that capital is flowing to the "brain" of the robot — the model layer — rather than just the hardware. The company's emphasis on proprietary data collection hardware and data loop closure indicates a belief that high-quality, real-world data is the key differentiator in building generalizable robot foundation models.
The timing is notable: while many humanoid robot startups are still focused on hardware iteration and proof-of-concept demos, Deta is betting that the software and data infrastructure will determine long-term winners. The company's strategy mirrors the early days of autonomous driving, where companies that invested early in data pipelines and simulation eventually pulled ahead.
However, the path to commercial deployment in industrial settings is long and capital-intensive. Deta will need to demonstrate that its foundation model can generalize across tasks and environments, and that its data collection approach scales cost-effectively. The Angel++ round gives it a runway to attempt this, but the company will likely need to show concrete industrial partnerships and deployment milestones in the next 12-18 months to secure follow-on funding at a higher valuation.
For startups watching this space, the key takeaway is that the market is beginning to differentiate between hardware plays and model-layer plays. Deta's raise suggests that investors are willing to pay a premium for the latter, especially if the company can demonstrate a clear path to data advantage and industrial applicability.
Source: 36氪.