General Intuition raises $320M to train agents on game data
General Intuition closed a $320 million Series A at a $2.3 billion post-money valuation on June 25, 2026, positioning itself as a key player in the emerging AI agent training market TechCrunch. The round was led by Khosla Ventures and included participation from General Catalyst, Hillspire, Bezos Expeditions (Jeff Bezos), venture investor Eric Schmidt, professional racing driver Nico Rosberg, and researchers from Google DeepMind and MIT LinkedIn.
The funding brings General Intuition's total disclosed funding to $454 million, including a $134 million launch round from October 2025. The company, which spun out of Medal—a platform founded by CEO Pim de Witte that allows gamers to upload and share video game clips—is using the capital to scale compute capacity through a partnership with CoreWeave, pre-train its next model version, and broaden API access by end of summer The Robot Report.
Training Agents on Gameplay Data
General Intuition's core thesis is that millions of hours of video game footage contains rich spatial-temporal reasoning patterns applicable to real-world robotics and autonomous systems. The company trains AI agents using gameplay data to develop capabilities relevant to physical robot control and drone navigation. This approach differs from traditional synthetic data generation by leveraging the complexity and diversity inherent in human player behavior across competitive and open-world game environments.
The strategic timing of this raise reflects accelerating enterprise interest in grounding AI agents in realistic sensorimotor learning. Game data provides naturally varied environments, physics interactions, and decision-making scenarios that transfer to real-world deployment contexts in logistics, manufacturing, and autonomous mobility.
Capital Deployment and Near-Term Milestones
The Series A proceeds will fund three primary initiatives: expanding computational infrastructure partnerships to support model training at scale; advancing the next generation of the company's pre-training framework; and opening its agent API to enterprise developers before the end of summer 2026. These moves position General Intuition to move from research validation to commercial adoption cycles.
The investor syndicate signals confidence in both the technical approach and market timing. Khosla's lead role underscores thesis-driven venture conviction in AI agent training methodologies. The participation of Bezos Expeditions and senior technical talent from frontier labs suggests institutional belief that game-trained agents represent a viable path to commercially viable robotics and autonomous systems.
With $454 million in total funding and a $2.3 billion valuation, General Intuition is now among the highest-capitalized AI agent training companies. The next critical milestone is demonstrating measurable transfer learning gains from game data to real-world robotic task completion metrics.
