Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
IndustryInsight · On RadarTNW→ original

Paris-Based Genesis AI Raises $500M at $3.5B Valuation for Eno Robots

Paris startup Genesis AI has raised $500M at a $3.5B valuation to bring its Eno robots to factories and warehouses. The robot learns from people: more than 100 workers wear sensor gloves and head-mounted cameras, and their movements are transferred to the robot's hands. So far, 30–40 machines have been built, and US series production is set to begin next year. The takeaway: investors are betting on a general-purpose model, but the bottleneck is hardware wear. The round was led by Khosla Ventures, Premji Invest and Eclipse. The robot's hands withstand only a few weeks of normal work. A pilot with LG CNS is running in production and logistics, with broader rollout expected in 2027.

AI-processed from TNW; edited by Hamidun News
Paris-Based Genesis AI Raises $500M at $3.5B Valuation for Eno Robots
Source: TNW. Collage: Hamidun News.
Listen to article

The market for "physical AI" has landed another large check. Paris-based startup Genesis AI announced a $500M Series A at a $3.5B valuation. The money is meant to build and deploy robots that will first work in factories and warehouses. The company is only two years old, yet it already builds its own model, its own hardware and its own simulator. For anyone tracking robotics, this is a signal: investors are willing to pay for a general-purpose robot before it has become reliable.

Who invested and what the money will fund

The round was led by Khosla Ventures, Premji Invest and Eclipse. NVentures from Nvidia, AMD Ventures, Bpifrance, HSG, Samsung Next and Ericsson Ventures also participated. Angel investors include former Google CEO Eric Schmidt, Xavier Niel and former Tesla CFO Zach Kirkhorn.

The company was founded in 2024 by Zhou Xian, who became CEO, and Théophile Gervet, a former researcher at Mistral AI. In July 2025 they raised a $105M seed round. The new funds will go toward producing Eno robots, expanding data collection and training models. According to Zhou, the goal is "the best general-purpose robot that works reliably in the real world, day after day."

How the stack works: gloves, model and simulator

The core idea behind Genesis is in how the robot learns. Its GENE model is trained on recordings of people working in sensor gloves. According to the company, the gloves map one-to-one onto the robot's hands, so human movements don't need complicated conversion. It's similar to an apprentice watching a master, except the "eyes" are sensors on the hands and a camera on the head. Right now, more than 100 workers around the world wear the gloves and cameras.

The second part is the Genesis World simulator. The model runs through millions of scenarios there before it goes into a real warehouse. Developers will recognize the approach: it's like testing code against synthetic loads before production, except the cost of an error is higher, because it means broken boxes and fingers.

Eno itself is two arms on a rotating wheeled base. It has no head. Zhou explained that he doesn't see why a robot needs a head, since the "brain" isn't there anyway.

What already exists and where the weak points are

For now, the scale is more modest than the valuation suggests. Genesis has built about 30–40 Eno robots and is testing them with several customers whose names it has not disclosed. A contract manufacturer will begin mass production in the US next year. Among the named partners is South Korea's LG CNS, with which Eno is being tested in production and logistics. The company expects broader deployment in 2027.

The founder openly names the main constraint: the hardware. Under normal work, the robot's hands last only a few weeks before they need replacing. For a warehouse, where a machine has to operate in shifts, this is a serious problem, because repairs and replacements eat into the economics.

A demonstration for TechCrunch is also telling. Genesis's model controlled a pair of standard off-the-shelf manipulators. On the first try they packed a box and applied a label; on the second, they stumbled. This is a typical problem in modern robotics: a scenario works in a demo, but consistency from run to run remains the main challenge.

Zhou promised that next year a robotics model will appear that "just works out of the box." This is a forecast, not a result, and it should be read with allowance for founders' familiar optimism.

What this means for the industry

First, large sums are flowing not into "yet another humanoid" but into a combination of data, simulator and hardware for specific tasks: factories and warehouses. Second, the bet on collecting data from people wearing gloves suggests that data is regarded as as scarce as computing. Third, the involvement of Nvidia, AMD and Samsung shows chipmakers' interest in getting models out beyond the screen.

For product managers and founders, the practical takeaway is simple: evaluate not the demo but the uptime without intervention and the cost of maintenance. Genesis has money, a team and partners for now, but there is no confirmation of production reliability yet. The next milestones are series production in the US and the scale of pilots in 2027.

Story

Key Questions Answered

Who invested and what the money will fund?

The round was led by Khosla Ventures, Premji Invest and Eclipse. NVentures from Nvidia, AMD Ventures, Bpifrance, HSG, Samsung Next and Ericsson Ventures also participated. Angel investors include former Google CEO Eric Schmidt, Xavier Niel and former Tesla CFO Zach Kirkhorn. The company was founded in 2024 by Zhou Xian, who became CEO, and Théophile Gervet, a former researcher at Mistral AI. In July 2025 they raised a $105M seed round.

How the stack works: gloves, model and simulator?

The core idea behind Genesis is in how the robot learns. Its GENE model is trained on recordings of people working in sensor gloves. According to the company, the gloves map one-to-one onto the robot's hands, so human movements don't need complicated conversion. It's similar to an apprentice watching a master, except the "eyes" are sensors on the hands and a camera on the head. Right now, more than 100 workers around the world wear the gloves and cameras. The second part is the Genesis World simulator.

What already exists and where the weak points are?

For now, the scale is more modest than the valuation suggests. Genesis has built about 30–40 Eno robots and is testing them with several customers whose names it has not disclosed. A contract manufacturer will begin mass production in the US next year. Among the named partners is South Korea's LG CNS, with which Eno is being tested in production and logistics. The company expects broader deployment in 2027. The founder openly names the main constraint: the hardware.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…