General Intuition Raises $320 Million at $2.3 Billion Valuation for Agentic AI

3 articles · Updated · TechCrunch · Jun 25

$320 million in new funding values General Intuition at $2.3 billion, lifting its total disclosed funding to $454 million after launching last October with a $134 million round.
Medal’s gameplay archive underpins the pitch: the startup says millions of hours of clips with button-press action labels help train one model that can move from Fortnite play to simulated worlds and quadruped robots.
Khosla Ventures led the round, joined by General Catalyst, Jeff Bezos, Eric Schmidt, Nico Rosberg, and researchers at Google DeepMind and MIT, betting gameplay data can reduce the need for costly real-world robotics training.
Most of the cash will fund compute through CoreWeave and pre-training the next model, while a broader API rollout is planned by the end of summer for customers in gaming, simulation, and robotics.
The company is positioning itself as a model provider rather than an acquirer target or defense contractor, though investors and executives acknowledge simulation-to-real-world performance at scale remains unproven.

Sources

Left100%

TechCrunch5h ago

General Intuition Raises $320M at $2.3B Valuation for Agentic AI

gamesbeat.com5h ago

General Intuition raises $320M at $2.3B valuation for AI frontier models based on gameplay | exclusive interview - GamesBeat

TechCrunch4h ago

General Intuition Raises $320M at $2.3B Valuation for AI Agent Model

Is an AI learning from Fortnite developing true intuition or just becoming the ultimate video game mimic?

Can an AI trained on video games build a lasting moat against tech giants with vastly more resources?

How will an AI brain trained on virtual physics adapt to the unscripted chaos of the real world?

General Intuition Raises $320M at $2.3B Valuation to Lead Agentic AI with Proprietary World Models and Data Moat

Overview

General Intuition has secured $320 million in funding, raising its valuation to $2.3 billion and highlighting strong investor confidence in its innovative approach to AI. Unlike others, the company builds world models not to sell, but to train its own advanced agents, making these agents its main product and a key market differentiator. This strategy is powered by a proprietary and highly diverse dataset, including Medal's clips, which features a wide range of physics engines, action spaces, and unique interaction data. This unique data foundation underpins General Intuition’s viability and ambitious growth in the AI industry.

...