Generalist said it has raised $400 million in fresh funding, a round that values the robotics startup at $2 billion. The company said the new capital will support its effort to build what it calls physical AGI, or general intelligence designed to operate in the real world.
The funding brings Generalist’s total capital raised to more than half a billion dollars. The company framed the round as backing for a broader push to develop robot intelligence that can work across environments and hardware types, from factory machinery to home robots and systems used in space.
Radical Ventures led the round, with participation from 8VC, Union Square Ventures, Hanabi Capital and Norwest. Existing investors also took part, including NVIDIA, Boldstart Ventures, Spark Capital, Bezos Expeditions and NFDG. Among the new angel investors were Bin Lin, Fei-Fei Li and Naval Ravikant.
Generalist has tried to position itself as a research-heavy robotics company focused on scaling model performance through real-world experience. In its announcement, the startup pointed to two recent systems, GEN-0 and GEN-1, as evidence that its approach is progressing toward commercial use.
According to the company, GEN-0 showed that robotics models trained on large-scale physical data can follow scaling laws similar to those seen in other areas of AI. Generalist said the result suggested that larger models and more experience in the physical world could lead to more capable robots.
The company said GEN-1, introduced in April, moved the technology closer to practical deployment. It described the system as reaching 99% reliability on a range of tasks, operating up to three times faster than earlier state-of-the-art systems and learning new physical skills. Generalist also said the model showed signs of improvisational problem-solving.
Generalist said the funding will help it continue building newer models, expand its physical data systems and grow its compute and training infrastructure. It also plans to work more closely with industries that could use the technology in everyday settings.
The company stressed that its goal is not tied to any one robot type or method. Instead, it said the important layer is intelligence that can move between different form factors and settings. That could include humanoid robots in homes, robotic arms in factories, mobile machines in warehouses and autonomous systems in space.
Generalist argued that robotics is still early in its development, but said the field could benefit from a cycle in which better models produce more useful work, which in turn generates more data for the next generation of systems.
The startup said it sees physical AI as a path toward broader machine intelligence, developed through systems that learn by acting in the world and improve through experience.
The latest financing underscores continuing investor interest in robotics and AI systems that can extend beyond screen-based software into physical tasks. It also gives Generalist more resources as it tries to turn recent technical progress into products that can be deployed commercially.