Robotics Startups Attract Record Venture Funding as China and US Race for Physical AI Dominance

Robotics startups have experienced explosive growth in venture capital this year, with investment concentrated heavily in China and the United States, according to Crunchbase data. During the third quarter of 2026, global venture funding to robotics startups reached $9.9 billion, representing a more than twofold increase from the $4.3 billion deployed in Q3 2025. The sector has maintained momentum throughout the year, with funding in the first three quarters of 2026 climbing approximately 185 percent compared to the same period in 2025.

China and the US lead

China has captured the largest share of robotics investment this year with $13.2 billion, followed closely by the United States at $12.1 billion. Germany ranks third among funding destinations, though significantly behind with $1.7 billion invested in robotics startups.

Within China, the most substantial funding rounds have gone to Guangzhou-based Xpeng Robotics, which secured $900 million for its work developing AI models and physical robots, and Shenzhen-based AI² Robotics, which closed a $736 million Series C round to advance intelligent robots for manufacturing, healthcare and service applications.

The United States has also seen billion-dollar funding events this year. Travis Kalanick's Los Angeles-based Atoms, which operates a ghost kitchen business alongside robotics divisions focused on mining and transport, raised substantial capital. Pittsburgh-based Skild AI secured $1.4 billion to develop Skild Brain, its robotics foundation model.

Germany-based Neura Robotics completed a notable $1.4 billion funding round for its cognitive robots, which integrate AI and data infrastructure to build machine intelligence.

Building out the data and model layer

The infrastructure supporting physical AI systems is expanding rapidly, though the challenges differ substantially from those faced by language models. Ulrik Stig Hansen, co-CEO of San Francisco-based Encord, a provider of the data layer for physical AI, explained the distinction at Crunchbase News's inaugural Long Horizon Physical AI Summit in San Francisco. "We're just seeing physical AI take off," Hansen told Crunchbase News.

While compute and storage capabilities are being developed to support these systems, physical world applications require significantly more time to mature than large language models trained on internet-scale text data. According to Hansen, "Physical AI is very different because there is no internet of physical AI data, and so the data first has to be collected, and you have to convert it from sensor tokens to action tokens."

Physical AI systems demand substantially larger datasets than their digital counterparts, incorporating multimodal data that combines vision, tactile, force and torque information. Hansen noted that "Storing all that data, curating all that data, enriching all that data, and validating that all the data works for these systems is a very expensive operation."

The sector currently remains in the pre-training phase of physical AI development, with the post-training phase expected to arrive in the near term, according to Hansen.

More winners

Approximately 75 robotics companies have achieved unicorn status and raised funding within the past three years. Of these, more than half—45 companies—reached unicorn valuation during 2026 alone. Among the most highly valued robotics startups, competing firms are pursuing data and model development alongside diverse vertical applications spanning heavy industry, manufacturing, logistics and household use cases.

While robotics funding continues to accelerate and new unicorns emerge regularly, the sector's ultimate trajectory remains unclear as companies pursue further technological advances. Hansen observed that "There are a lot of companies that are raising capital, and I think a lot of them are going to win, just because there's so many different types of applications to the physical world."