Genesis AI has raised $500 million in a Series A that values the robotics company at $3.5 billion once the new money is counted. Genesis AI announced the round on November 17, 2026, 16 months after a $105 million seed in July 2025.
- Disclosed funding before this round: $105 million
- Disclosed funding after it: $605 million ($105 million + $500 million)
Take $500 million off the $3.5 billion and you get a $3.0 billion price before the cash arrived. New investors hold about 14.3% of the company ($500 million ÷ $3.5 billion), so everyone who owned shares on November 16 now owns 14.3% less of it. For scale, the cheque is 4.8 times the seed ($500 million ÷ $105 million) and matches the $500 million Torus raised for its AI computing business.
Premji Invest, Khosla Ventures and Eclipse co-led. Nine more institutions joined, Nvidia, AMD Ventures, Bpifrance, Zoom and Samsung Next among them, plus five individuals including Eric Schmidt and former Tesla CFO Zach Kirkhorn.
The company lists four uses for the money: more training data, better hand manipulation, putting Eno to work with customers, and hiring researchers and engineers. It gave no headcount target.
Customers come in a fixed order, with factories, warehouses and laboratories first, hospitality and healthcare second, and consumer use after that. A contract manufacturer starts mass production in the US in 2027, TechCrunch reported, and a handful of unnamed customers are running pilots now.
What does $3.5 billion buy when only 30 to 40 robots exist?
Genesis has built 30 to 40 Enos so far. Divide the valuation by the larger count and each finished robot carries $87.5 million of it ($3.5 billion ÷ 40). That tells you the 12 institutions in this round are paying for the 2027 production line and the training data, since 40 machines can’t justify the price on their own. A Series A normally funds something much smaller: this one is 16.7 times the $30 million Series A that SE3 raised from Accel ($500 million ÷ $30 million).
The data side has one hard number, more than 100 workers wearing sensor gloves and headset cameras. The hardware side has a weaker one. CEO Zhou Xian told reporters the robot’s hands last only a few weeks of regular work before they need replacing. In a press demo the model ran on off-the-shelf arms because Eno wasn’t ready, packed and labelled a box on the first run, then fumbled the second.
What the figures leave out matters more to you than what they include. There is no production volume for 2027, no price per Eno and not one named customer. Three things will show whether the $3.5 billion holds up: the first unit count from the contract manufacturer in 2027, the first customer willing to be named, and whether hand life moves from weeks to months.
