Genesis AI Raises $500M Series A to Build Human-Handed General-Purpose Robots


Robotics investors are shifting their attention from impressive demos to reliability on the factory floor. A robot that nails a task once is not the same as one that does it right every day, and Genesis AI has just raised $500 million on the bet that closing that gap means building the model, the hand, the robot and the simulator in-house.

Genesis AI co-founders

Genesis AI is building a general-purpose robot that can learn any kind of physical work. The system has four parts: GENE, a robot foundation model; a robotic hand sized and shaped like a human hand, paired with data-collection gloves that map one-to-one onto it; Genesis World, a simulator built to match reality; and Eno, its first robot, which combines a compact mobile base with precise dual-arm manipulation. Touch and motion data from people wearing the gloves trains the model, a smarter model makes the robot more capable, and robots in the field generate more data, a loop the company says speeds up every step.

Reliability is the reason Genesis insists on owning the full stack. By its math, a robot that succeeds 99% of the time at each step has only an 82% chance of finishing a 20-step workflow without error. If model, hardware and deployment are built separately, the weakest link sets the ceiling, so the company argues it has to control all three to fix problems at their source. In its design, field failures become evaluation scenarios in the simulator, new training data for the model and design input for the next robot.

The company was founded by CEO Zhou Xian and President Théophile Gervet, who, according to press reports, met in a lab at Carnegie Mellon University; Gervet previously worked as a research scientist at Mistral AI. Genesis grew out of an open robotics-simulation project involving researchers from many universities and came out of stealth in July 2025. It has offices in Paris and San Carlos, California, and close to 100 employees.

Genesis announced a $500 million Series A. The company’s announcement does not state a valuation; press reports put it at $3.5 billion. Premji Invest, Khosla Ventures and Eclipse co-led the round, with participation from T. Rowe Price, Durable Capital Partners, Bpifrance, HSG, NVIDIA, AMD Ventures, Zoom, Samsung Next and Ericsson. Individual backers include former Google CEO Eric Schmidt, Zoom CEO Eric Yuan, Yahoo co-founder Jerry Yang, Xavier Niel and former Tesla CFO Zach Kirkhorn. Disclosed funding now totals $605 million.

The round follows a $105 million seed in July 2025 co-led by Eclipse and Khosla Ventures. The new capital will fund robot development and production, data collection and hiring. Genesis plans to start in manufacturing and logistics, then move into services such as hospitals and hotels, and later homes and outdoor settings.

Field validation is only beginning. According to press reports, 30 to 40 Eno units have been built and are being piloted with a handful of customers, and Korea’s LG CNS is testing the robot in manufacturing and logistics. The mass-production design is final, with a US contract manufacturer starting next year. TechCrunch noted that durability remains a challenge, with the robot hands needing replacement after a few weeks of regular work, and that in a demo the model packed and labeled a box on its first run but stumbled on the second.

Competitive Landscape

Robot-AI companies split into those that build only the brain and those that build the body too. Physical Intelligence raised $600 million at a $5.6 billion valuation in November 2025, Skild AI raised $1.4 billion, and Generalist AI raised a $400 million Series A at a $2 billion valuation. All three lean toward the model side.

Others build the hardware as well. Figure AI raised $1 billion at a $39 billion valuation, and Apptronik raised a $935 million Series A at a $5.3 billion valuation. Genesis sets itself apart by training its model directly on glove data captured with human-sized hands.

For more on the competitive landscape, see here.

Share

Leave a Reply

Your email address will not be published. Required fields are marked *