Simulation startup Antioch raises $8.5M to build 'Cursor for physical AI'

Simulation startup Antioch has raised $8.5 million in seed funding to build tools that help robot developers create and test autonomous systems in virtual environments. The round values the New York-based company at $60 million. Led by venture firms A* and Category Ventures, the investment aims to address what robotics engineers call the "sim-to-real gap". the challenge of making virtual simulations realistic enough that robots trained in them can operate reliably in the physical world. Antioch allows developers to create digital replicas of robotic hardware and connect them to simulated sensors that mimic real-world data. The platform enables testing of edge cases, reinforcement learning, and generation of training data without requiring expensive physical testing facilities. "How can we do the best possible job reducing that gap, to make simulation feel just like the real world from the perspective of your autonomous system?" said Antioch cofounder Harry Mellsop. The company compares its product to Cursor, the AI-powered software development tool, but for physical AI systems. Antioch builds on simulation models from Nvidia, World Labs, and others, adding domain-specific libraries to make them accessible to robotics developers. "What happened with software engineering and LLMs is just starting to happen with physical AI," said Çağla Kaymaz, a partner at Category Ventures. "With software, you can have bad coding tools, and the risk is generally contained to the digital world. In the physical world, the stakes are much higher." Antioch's initial focus is on sensor and perception systems for automated vehicles, construction machinery, and aerial drones. While targeting startups, the company has also engaged with large multinational corporations already investing heavily in robotics. Former Cruise executive and Foxglove founder Adrian Macneil, an angel investor in Antioch, emphasized simulation's importance for safety-critical applications. "It's not possible to drive enough miles in the real world," he said at a recent conference. Researchers at MIT's Computer Science and Artificial Intelligence Laboratory are already using Antioch's platform to evaluate large language models by having them design robots and test them in simulated environments. The funding comes as the robotics industry seeks more scalable alternatives to physical testing, which often requires building mock facilities or collecting extensive real-world data. Antioch aims to provide the simulation infrastructure needed to accelerate development of autonomous systems across multiple industries.