French robotics-focused startup Genesis AI has released its first foundational model, GENE-26.5, along with a human-like hand built in-house that is capable of carrying out complex tasks such as cooking or solving a Rubik's Cube. The company, which completed a $105 million (£77.2 million) seed funding round in July of last year, began with a focus on a model to power real-world robotics, then built its own hand in order to have more accurate control over the full stack, said co-founder and chief executive Zhou Xian.

Unlike robotic hands built by dozens of other companies, which tend to have two or three fingers, the unit from Genesis AI closely resembles a human appendage, allowing data to be more directly transferred to the device. This anthropomorphic approach is a deliberate strategy to bridge the gap between simulated training data and real-world manipulation. By mimicking the human hand’s anatomy—including 23 degrees of freedom and tactile sensors—Genesis aims to accelerate the learning curve for AI models tasked with dexterous operations. The hand’s design also facilitates easier teleoperation, where a human operator wearing a sensor glove can control the robot hand in near-real-time, capturing nuanced movements for training data.

AI Model and Training

GENE-26.5, the company’s flagship AI model, is a large-scale neural network trained on a diverse dataset that includes both simulated and real-world manipulation tasks. The model uses a transformer architecture optimized for robotic control, allowing it to generalize across a wide range of actions—from grasping objects with precision to performing multi-step sequences like assembling parts or preparing food. Genesis claims that GENE-26.5 achieves state-of-the-art performance on several benchmarks, including the famous Rubik’s Cube solving task, which requires fine motor control and problem-solving abilities.

The decision to build the hand in-house was driven by the need for end-to-end vertical integration. “By controlling both the hardware and the software, we can iterate faster and ensure that the model’s outputs translate seamlessly into physical actions,” Zhou Xian explained in a recent interview. This strategy mirrors that of other leading robotics firms like Tesla and Boston Dynamics, but with a focus on small-scale precision rather than large-scale locomotion.

Sensor Gloves and Data Collection

Genesis has also created lightweight sensor gloves that can be worn by workers in fields such as pharmaceuticals or manufacturing, capturing data that can be passed along to the AI model – although whether workers would be happy to help train a potential robotic replacement remains in question. Data from the gloves can be combined with information gleaned from videos of workers carrying out tasks, Genesis said. The gloves are embedded with inertial measurement units (IMUs) and strain gauges that record hand and finger movements at high resolution. This multimodal data, when fed into the AI model, helps the system understand not just the kinematics of a task but also the forces required—such as how hard to grip a delicate vial or how to apply torque when turning a screwdriver.

In a demonstration video, Genesis showed a pair of its hands carrying out complex or delicate processes such as preparing a smoothie or playing the piano. The smoothie demonstration involved multiple steps: picking up a blender jar, adding ingredients from various containers, securing the lid, pressing the blend button, and then pouring the finished drink into a glass. Each action required precise coordination between the two hands and real-time adjustments based on visual feedback. The piano playing, while not concert-level, showed the hand’s ability to depress keys with varying force, a sign of its fine motor control.

Funding and Team

Co-founder and president Theophile Gervet previously worked for France’s Mistral, and the company retains a substantial presence in Europe that it attributed to the continent’s large talent base and potential market of industrial customers. The company said its team of 60 people is split roughly evenly between the US and Europe, with a slightly larger group in the US base in California. In addition to San Carlos, California, it has offices in Paris and London.

The seed funding round last year was co-led by Eclipse and Khosla Ventures, with additional backing from Bpifrance, HSG, former Google chief Eric Schmidt, telecoms entrepreneur Xavier Niel, MIT Computer Science and Artificial Intelligence Laboratory director Daniela Rus and Apple distinguished scientist Vladlen Koltun. This constellation of investors reflects the high level of interest in embodied AI and the convergence of robotics with large language models. The $105 million round is one of the largest seed rounds ever for a European robotics startup, signaling strong confidence in Genesis’s approach.

Future Ambitions

The company said it is working to supplement the mechanical hand with a full-body, general purpose robot. While details are scarce, the goal is to create a humanoid robot that can operate in unstructured environments—such as homes, factories, and hospitals—performing tasks that currently require human dexterity and adaptability. Genesis faces stiff competition from other startups and tech giants developing general-purpose robots, including Figure AI, 1X Technologies, and Tesla’s Optimus. However, Genesis’s focus on the hand as a primary differentiator may give it an edge in tasks that require a high degree of manual skill.

The broader robotics industry has seen a surge of investment in recent years, driven by advances in AI and the need for automation in labor-shortage sectors. According to a recent report by the International Federation of Robotics, global robot installations reached a record high in 2024, with the service robotics segment growing particularly fast. Genesis aims to capture a slice of this market by offering both a hardware platform and a software model that can be licensed to other manufacturers. The company envisions a future where its AI model powers a variety of robot embodiments, from warehouse pickers to surgical assistants.

To support its growth, Genesis is actively recruiting engineers with expertise in machine learning, mechanical design, and control systems. The dual presence in the US and Europe allows it to tap into talent pools from both Silicon Valley and the Paris-Saclay research cluster. The European offices also benefit from proximity to industrial customers in the automotive, aerospace, and pharmaceutical sectors, which are increasingly looking to deploy advanced robotics for complex assembly and packaging tasks.

Meanwhile, the ethical implications of such technology continue to be debated. The sensor gloves, while useful for training, raise concerns about worker surveillance and job displacement. Genesis acknowledges these issues but argues that its technology can augment human workers rather than replace them entirely, at least in the near term. “We see the gloves as a tool for collaboration, not control,” said Theophile Gervet. “By capturing expert knowledge, we can make robots better assistants, allowing humans to focus on higher-level decision-making.”

In the coming months, Genesis plans to release more details about its full-body robot and additional capabilities of the GENE model. The company also intends to open-source parts of its dataset and simulation environment to foster research in dexterous manipulation, following a trend set by Meta and Google. If successful, Genesis’s human-like hand and AI model could become key building blocks for the next generation of general-purpose robots.

Source: Silicon UK News