Flapping Airplanes, a San Francisco AI research lab founded in late 2025, has raised $1B at a $6B valuation, Bloomberg reported.
Index Ventures and Lux Capital led the round. The lab's valuation has quadrupled since January. By simple division, that puts the January figure near $1.5B, though that is our arithmetic and not a number Bloomberg gave.
The lab has no product. Its research focuses on training AI models with less data and less compute, the industry's shorthand for raw computing power.
That focus explains the interest. Data and compute are the two largest inputs to building modern AI models, which is why the leading labs spend so heavily on chips and data centers. A group that gets comparable results with less of either would change the cost structure of the field. That is the pitch, and it is a research bet. Investors are paying for a result that has not yet shipped in any form the public can test.
The pace is the other notable detail. A company less than a year old has seen its valuation grow fourfold and has pulled in $1B before releasing anything. The open question is where an efficiency gain would end up if it works: as an edge the lab keeps, or as a technique the larger labs, with far more compute on hand, absorb and scale up first.