Genesis AI — Startup Profile

Robot intelligence trained with simulated and real-world experience

HQ: Palo Alto, California, and Paris, France | Founded: 2024 | Employees: Not disclosed in the reviewed sources | Stage: $105M seed announced July 2025; later financing talks not treated as closed | Website: https://www.genesis.ai/

About

Zhou Xian and Théophile Gervet founded the business in December 2024, according to TechCrunch. Xian's background includes a Carnegie Mellon robotics PhD; Gervet previously worked at Mistral and Skild AI. The company initially emphasized an in-house simulation engine and a horizontal robotics foundation model. The Robot Report identifies Paris and Palo Alto offices and a dual pipeline of synthetic and real data.

Its current site highlights Eno and a customer waitlist. This is a product-development signal, not evidence of general availability or completed deployments. July 2026 reports of a potential larger financing are not included as funded rounds because closing was not established. The academic/open-source Genesis ecosystem should not automatically be equated with the company's proprietary IP or commercial product.

The Take

Genesis AI develops general-purpose robotics models and a data engine combining physics simulation with real-world robot experience. It emerged from stealth with $105M in July 2025, co-led by Eclipse and Khosla Ventures. Current official pages invite interest in Eno and say targeted deployments are planned by the end of 2026, not already completed.

Editorial assessment: owning simulation and real-data collection could reduce the cost of training diverse skills. The important proof is whether the synthetic experience transfers to reliable physical tasks and whether deployment economics beat task-specific automation.

Evidence boundary: financing announcements establish disclosed transactions, not audited operating performance. Unavailable headcount, margins, and retention are not estimated, and company claims remain attributed.

Key Metrics

Founded: 2024
Location: Palo Alto, California, and Paris, France
Reviewed financing: $105M — Seed (2025-07-01)
Product: Robotics foundation model and data engine
Availability: research and development

Funding Rounds

Round: $105M (2025-07-01) — Eclipse and Khosla Ventures

Key People

Zhou Xian
Théophile Gervet

Key Competitors

Physical Intelligence —
Generalist AI —

Timeline Highlights

2025-07-01: Seed: $105M

Similar Startups

Physical Intelligence — Building foundation models to power any robot in the physical world
Generalist AI — Embodied foundation models for robots across tasks and form factors
Skild AI — The Skild Brain — a unified robotics foundation model for any robot, any task
Agility Robotics — Humanoid robots for work in warehouses and factories.
Bright Machines — Software-defined automation for electronics manufacturing.

More Robotics & Automation Startups

Apptronik — Apollo — the humanoid robot built for work
Agility Robotics — Humanoid robots for work in warehouses and factories.
Dexterity — Robots that handle boxes and pallets in logistics facilities.
Gecko Robotics — Robots that inspect critical industrial infrastructure.
Chef Robotics — Robots that help prepare food at scale.

Category: Robotics & Automation Startups