Robotics startup Generalist has reached a $3 billion valuation after raising nearly $200 million in additional capital led by venture firm 8VC, according to two people with knowledge of the deal and a regulatory filing reviewed by TechCrunch. The fresh funding is an extension of a $400 million Series B led by Radical Ventures that the company announced in June at a $2 billion valuation. The new capital brings the round’s total to $600 million, making Generalist one of the most heavily funded robotics AI companies in the world.
Generalist was founded in 2024 by former Google DeepMind researchers Pete Florence and Andy Zeng, along with former Boston Dynamics engineer Andrew Barry. The San Mateo-based startup builds AI foundation models designed to work across a wide range of robots, rather than building the robots themselves. Its latest model, Gen 1.5, enables robots to master new tasks from video demonstrations as short as 3 to 12 seconds, a capability the company says marks a shift toward practical, deployable physical AI.
Investors Race to Back the Robotics Brain
The funding surge reflects a growing bet among venture investors that robotics may soon reach its own ChatGPT moment, where machines can perform general tasks without being explicitly trained for each one. Generalist’s approach positions it as the intelligence layer sitting on top of whatever hardware a customer already owns, rather than competing in the crowded humanoid robot hardware space.
Investors are pouring capital into this thesis at record pace. Physical Intelligence, which builds similar foundation models for robots, is reportedly valued at $11 billion. SoftBank-backed Skild AI sits at $14 billion. Genesis AI was in talks last month to raise at a $3 billion valuation as well. Together, these companies represent billions of dollars in venture funding targeting what NVIDIA CEO Jensen Huang has described as a potential trillion-dollar industry.
The competitive landscape is intensifying rapidly. Generalist’s model competes directly with Physical Intelligence and Skild AI on the software side, while both companies pursue broadly similar goals of creating a universal brain for machines. What sets Generalist apart is its hardware-agnostic approach, running on off-the-shelf robot arms with standard pincer grippers rather than requiring specialized equipment or purpose-built humanoid platforms.
Generalist’s team brings deep technical credibility to the race. Pete Florence previously led work on RT-2 and PaLM-E at DeepMind, two of the most cited papers in robotic AI. Andy Zeng was a senior scientist at DeepMind, and Andrew Barry spent years at Boston Dynamics working on manipulation and locomotion systems. Their combined experience across research and hardware gives the startup an edge in bridging the gap between theoretical capability and real-world deployment.
The Training Data Problem That Defines the Race
One of the biggest challenges in building general-purpose robotics AI is the lack of training data. Unlike large language models, which can be trained on the vast text corpus of the internet, robots require physical interaction data that is expensive and time-consuming to collect. Generalist has attempted to solve this problem through proprietary Data Hands, wearable gripper devices that convert human hands into robot-style input devices for collecting training data at scale.
The company says it has amassed over 500,000 hours of demonstration data using this approach, which feeds into its foundation models. This data collection strategy has attracted early backing from a notable group of investors including NVIDIA, Bezos Expeditions, Union Square Ventures, and prominent AI researcher Fei-Fei Li. Angel investors also include Xiaomi co-founder Lin Bin and Zoom CEO Eric Yuan.
Generalist is currently working with a handful of customers to tailor its models for specific use cases, with early deployments focused on tasks that previously required human dexterity and judgment. The company is not alone in pursuing these deployments, but its focus on a software-only model that works across different hardware platforms could give it broader reach than competitors building their own robot bodies.
The race to build a universal robotics brain is still in its early stages, and some venture capitalists warn that a truly general model may still be years away. But the pace of funding suggests investors are unwilling to wait on the sidelines. Generalist’s leap from $2 billion to $3 billion in under two months is the clearest signal yet that the money is moving faster than the technology, and that investors believe the winner in physical AI will capture enormous value as robots move from factory floors into homes and offices.
The new capital will be used to accelerate research, expand the engineering team, and scale production. With $600 million in its latest round alone, Generalist now has the resources to compete with rivals valued at multiples of its own worth. The question is whether the startup can turn its technical lead into a product that redefines how robots learn, or whether the sheer volume of capital flowing into physical AI will fragment the market beyond any single winner.
discussion