Generalist AI's GEN-1 Robot Model Hits 99% Success Rate, Works on Any Hardware With Just One Hour of Training Data

Generalist AI's new GEN-1 robot model is achieving a 99% success rate on physical tasks like folding clothes and packaging items, up from 64%, while operating three times faster and requiring only one hour of training data per task on any hardware.

• Generalist AI debuts GEN-1, a hardware-agnostic robotics model achieving a 99% success rate on physical tasks like folding clothes and packaging items, up from a previous 64% success rate, while completing tasks three times faster.
• The model requires only one hour of robot data per task and is designed to work across any hardware form factor, with Generalist claiming it holds the world's largest pretraining dataset for robotics.
• Generalist AI, valued at $440 million, is working to close what researchers call the '100,000-year data gap' in physical AI by using pretraining to create a data flywheel that continuously makes robots smarter over time.