General Intuition Inc. announced a $220 million funding round on Sept. 29, 2026. The money comes from Valor Equity Partners, Atreides, Seven Seven Six, Point72, Khosla Ventures and General Catalyst. The round values the company at $6.2 billion.
The cash will be used to hire more AI researchers and to move the company’s newest algorithm, MIRA, from a research demo toward commercial use. MIRA is a "world model" that can generate synthetic video footage for training other AI systems, especially robots.
What makes MIRA different?
Most video‑generation AIs can only produce short clips before the quality collapses. MIRA claims it can run indefinitely without diverging, creating long, seamless simulations of fast‑moving objects.
It renders at 20 frames per second with a resolution of 720 × 576 pixels on a single Nvidia B200 graphics card. The efficiency comes from latent diffusion – a technique that works in a compressed "latent space" instead of raw video frames – and a relatively modest size of 5.6 billion parameters, far fewer than the trillion‑parameter models that dominate headlines.
Where could this matter to you?
If you work with robotics, simulation or video‑game development, MIRA could supply cheap, endless training data, cutting the time spent filming real‑world footage.
For everyday users, the technology may later appear as more realistic virtual environments, smarter game AI, or safer robots that have practiced countless scenarios in a simulated world before ever touching a physical object.
Why it matters
The funding lets General Intuition turn its demo into a usable tool for robot developers and simulators, potentially lowering the cost of training AI for real‑world tasks. For consumers, the tech could lead to richer virtual experiences and safer, better‑trained machines.
