Robotics Startup Genesis AI in Talks to Raise About $500 Million at $3 Billion Valuation

Genesis AI, a full-stack robotics startup, is in talks with investors to raise approximately $500 million at a $3 billion pre-money valuation, according to a Bloomberg report published July 23, 2026. The round would mark a dramatic step up for the company, which raised a $105 million seed round just over a year earlier.

Genesis AI builds foundational AI models for robotics and has expanded into hardware, including a human-like robotic hand developed in partnership with Chinese hardware company Wuji Tech. The company's stated mission, according to co-founder Zhou Xian, is to build the most capable robotic system, an ambition that pushed the team from a pure AI-model focus into building physical hardware as well.

What Happened: Genesis AI's $500 Million Funding Talks

Bloomberg reported that Genesis AI is in discussions to raise around $500 million in a new funding round that would value the company at approximately $3 billion before the new money is added. The talks are still ongoing and the final amounts could change, but the reported figures represent a roughly 28-fold increase in valuation compared to where Genesis AI stood after its 2025 seed round.

Genesis AI's previous funding came from a $105 million seed round closed on July 1, 2025, co-led by Eclipse and Khosla Ventures. That round drew a notable group of individual investors alongside the institutional backers, including Alphabet's former CEO Eric Schmidt, French billionaire Xavier Niel, robotics researcher Daniela Rus, and AI researcher Vladlen Koltun.

Key Details

Genesis AI is led by CEO Zhou Xian and President Théophile Gervet, who previously worked as a research scientist at Mistral AI. The company employs roughly 60 people spread across Paris, California, and London, with 40 to 45 percent of the team based in Europe and 50 to 55 percent in the United States.

The company's first model, GENE-26.5, was introduced in May 2026. Rather than stopping at software, Genesis AI built a human-like robotic hand in partnership with Wuji Tech, designed to reduce what the company calls the embodiment gap between how AI models are trained and how physical robots actually interact with the world. Genesis AI also built a sensor-loaded data collection glove that mimics the robotic hand, letting the company gather training data directly from human hand movements.

Why It Matters

A jump from a $105 million seed round to talks for a $500 million raise at a $3 billion valuation in about a year signals unusually strong investor confidence in Genesis AI's full-stack approach, building both the AI models and the physical hardware that runs them, rather than licensing models to existing robotics manufacturers.

That full-stack strategy sets Genesis AI apart from other well-funded AI infrastructure startups, such as Prime Intellect, that focus purely on software. By controlling both the model and the hand hardware it runs on, Genesis AI can tune training data collection, simulation, and physical design together, an approach the company argues produces more capable robotic systems than treating AI and hardware as separate problems.

Industry Context

Genesis AI's fundraising talks arrive amid intense investor interest in what the industry calls physical AI, AI systems designed to operate in and manipulate the physical world rather than purely digital environments. Genesis AI's demo of GENE-26.5-powered hardware reportedly showed the robotic hand performing tasks including cracking eggs, slicing tomatoes, preparing smoothies, playing piano, solving a Rubik's cube, and completing lab work, a broad range intended to demonstrate general-purpose dexterity rather than a single narrow task.

The company also built a proprietary simulation system to accelerate model training, a common approach among robotics AI companies seeking to generate large amounts of training data without the cost and time of exclusively physical, real-world testing.

What It Means for Users and the Industry

For the robotics industry, Genesis AI's rapid valuation growth, following a similar trajectory to Quantum Systems' $1.2 billion drone technology raise, adds to a wave of capital flowing into physical AI and humanoid or human-like robotics companies in 2026. A successful close near $3 billion would place Genesis AI among the better-funded startups in the category, giving it resources to scale hardware production and expand its data collection efforts beyond the current 60-person team.

For potential commercial partners and customers, from manufacturing to logistics, a well-capitalized Genesis AI could accelerate the timeline for deploying general-purpose robotic hands capable of the kind of varied, fine-motor tasks shown in its demos, though the company has not detailed a commercial product timeline beyond the GENE-26.5 research model.

What Happens Next

The funding talks remain in progress, and Genesis AI has not confirmed the final round size or valuation. If the reported $500 million raise closes near the discussed $3 billion valuation, it would be one of the larger funding events in the robotics AI category so far in 2026, and would likely accelerate the company's hiring and hardware production plans across its Paris, California, and London offices.

Final Takeaway

Genesis AI's reported talks to raise $500 million at a $3 billion valuation, up sharply from its $105 million seed round a year earlier, reflect how quickly capital is moving toward companies building both the AI models and the physical hardware for robotics. With GENE-26.5 already demonstrating a range of dexterous tasks and a full-stack strategy built around its Wuji Tech-developed robotic hand, Genesis AI is positioning itself as a leading bet in the physical AI category, assuming the round closes as reported.

Why Genesis AI Built Its Own Robotic Hand

Many AI-for-robotics companies focus exclusively on software, training models that other manufacturers' hardware can run. Genesis AI took a different path, building its own robotic hand in partnership with Wuji Tech specifically because the company concluded that AI models trained without matching, purpose-built hardware struggle to close what it calls the embodiment gap, the mismatch between how a model learns and the physical mechanics of the device meant to act on that learning.

The company's sensor-loaded data collection glove reflects the same logic. By capturing detailed data from human hand movements using a glove shaped to mirror the robotic hand's own design, Genesis AI can train its models on data that maps more directly onto the hardware it will eventually control, rather than adapting data collected for a generic or different physical form. That tight coupling between data collection, model training, and hardware design is central to the full-stack pitch the company is reportedly using to attract its latest round of investors.

FAQs