Genesis AI Unveils GENE-26.5 Robotic Brain With Human-Level Dexterity After $105 Million Seed Round

--- headline: "Genesis AI Unveils GENE-26.5 Robotic Brain With Human-Level Dexterity After $105 Million Seed Round" slug: genesis-ai-gene-265-robotics category: research story_number: "09" date: 2026-05-07 ---

# Genesis AI Unveils GENE-26.5 Robotic Brain With Human-Level Dexterity After $105 Million Seed Round

A robot that can crack an egg with one hand, solve a Rubik"s Cube in mid-air, and play a fast piano composition is no longer a research fantasy. Genesis AI, a Franco-American robotics startup barely a year old, on May 6 unveiled GENE-26.5, what it calls the first AI brain capable of giving robots human-level physical manipulation, alongside a proprietary robotic hand and data-collection system designed to make that intelligence scalable.

The announcement caps a rapid ascent for the San Carlos, California- and Paris-based company, which emerged from stealth in July 2025 with a $105 million seed round co-led by Eclipse and Khosla Ventures. Additional backers include Bpifrance, HSG, former Google CEO Eric Schmidt, French telecom billionaire Xavier Niel, MIT professor Daniela Rus, and AI researcher Vladlen Koltun.

Full-Stack Ambition

Where most robotics AI companies focus on either hardware or software, Genesis has pursued a full-stack approach, building the model, the hand, the data pipeline, and the simulation environment in-house.

"The model has always been the goal, because a better model means better intelligence," Genesis co-founder and CEO Zhou Xian told TechCrunch. "So we decided to go full stack."

GENE-26.5 is a foundation model trained on what the company describes as massive amounts of human-based internet videos, simulation data, and real-world demonstrations. To prove its capabilities, Genesis released a video showing its system performing a series of tasks that would challenge even a dexterous human: cooking a 20-step meal that includes chopping tomatoes and cracking eggs single-handed, preparing and serving a smoothie with coordinated two-hand control, conducting high-precision lab experiments with pipettes and liquid transfers, wire harnessing for electronics assembly, and solving a Rubik"s Cube with continuous in-air manipulation at speed.

Closing the Embodiment Gap

Central to the breakthrough is hardware that mirrors human anatomy. Genesis"s robotic hand matches the size and shape of a human hand, a deliberate departure from the two-finger grippers common across the industry. Paired with the hand is a lightweight data-collection glove embedded with tactile-sensing electronic skin.

"When worn by a human, the glove allows for a 1:1:1 mapping between the glove itself, the human"s hand and the robotic hand," the company said in its announcement. "This allows humans to seamlessly provide GENE-26.5 with high-quality data at scale that translates into robotic skills."

The glove, which Genesis says is 100 times cheaper than typical data-collection hardware and has demonstrated up to five times greater data-collection efficiency than traditional teleoperation methods, is designed to be worn during everyday work. The company is already engaging with industrial partners to deploy it in real-world environments, from pharmaceutical labs to manufacturing floors.

"That lets us collect a lot more data than was previously possible, to train a model that can do many more tasks," said Theophile Gervet, a former research scientist at Mistral AI who serves as Genesis"s president.

Data as the Moat

The data strategy extends beyond the glove. Genesis"s engine also taps egocentric video captured by workers wearing cameras, plus what the company characterizes as the world"s most advanced physics-based simulation platform. The simulation system allows teams to train and evaluate models far faster than physical testing, which Genesis notes is slow, expensive, and difficult to scale.

The company is collecting real-world data from tens of thousands of industrial workers using its sensor-equipped gloves, feeding that information back into its models. Customer engagements are expected to run three to five years, according to Vivian Sun, Genesis"s vice president of commercial and strategy, reflecting how deeply these systems integrate into industrial workflows.

Industry Context and Investment Significance

The $105 million seed round is among the largest in robotics history, and the investor roster signals broad confidence that general-purpose robotic intelligence is approaching commercial viability. "This marks an important milestone for their team and the robotics industry more broadly," said Eric Schmidt, who invested in the startup.

Genesis"s timing aligns with surging demand for industrial automation in Europe and beyond. German supplier Schaeffler said the same week that it expects its robotics order book to grow into the hundreds of millions of euros by 2030. The broader market for AI-powered robotics is projected to grow at a compound annual rate exceeding 25 percent through the end of the decade, driven by labor shortages, reshoring trends, and the falling cost of AI inference.

The company"s 60-person team is split roughly 40 to 45 percent in Europe and 50 to 55 percent in the United States, with offices in Paris, London, and Silicon Valley. Its founders, including Carnegie Mellon PhDs focused on AI and robotics, bring a mix of academic rigor and startup velocity that has attracted talent from organizations including Mistral AI.

What Comes Next

Genesis has said it will soon unveil its first general-purpose robot, a full-body system rather than just hands, that will serve as the culmination of the technology announced this week. Zhou told TechCrunch the broader roadmap remains unchanged: "Our goal is to build the most capable robotic system."

The open question is whether the company"s full-stack bet, controlling everything from the fingertip sensors to the foundation model, will prove to be its competitive moat or an overextension of resources. If GENE-26.5"s demo capabilities translate to real-world reliability, Genesis could set the pace for a new generation of general-purpose robots. If the gap between choreographed demos and messy factory floors proves wider than expected, the $105 million will have bought a very impressive science experiment. The next 12 months will tell.