Flapping Airplanes Raises $1B Series A at $6B Valuation to Build Data-Efficient AI

Flapping Airplanes, an AI research lab founded roughly a year ago, has raised a $1 billion Series A at a reported $6 billion valuation. Index Ventures and Lux Capital co-led the round, with participation from earlier backers Sequoia, GV and Kleiner Perkins, along with Valor, Atreides, A* and Nvidia, according to TechCrunch.

The valuation, first reported by Bloomberg, is roughly four times the $1.5 billion the company was valued at in its $180 million seed round about nine months ago.

What Flapping Airplanes Is Building

Flapping Airplanes is chasing a different scaling path than most frontier labs. Rather than training ever-larger models on ever-larger piles of internet text, the company wants to build AI systems that learn from far less data at far lower cost, or, in the founders' words, without "ingesting half the internet."

The team describes its technical foundation as "the world's most complete megakernel framework," which it says can overlap memory-bound inference with compute-bound training so GPUs are rarely idle. In plain terms, the lab is trying to squeeze more learning out of every chip-hour, a goal that matters as much to its own costs as to its research.

The lab was founded by Aidan Smith, a former founder and Thiel Fellow who worked at Neuralink in college; Asher Spector, a former statistics Ph.D. student at Stanford; and Ben Spector, a computer science Ph.D. student at Stanford who founded Prod while at MIT. Sequoia partner David Cahn has described the team as building the "young person's AGI lab."

How It Will Use the Funding

The company said the money will go toward hiring and pursuing its research goals. It has not disclosed product, revenue or user metrics.

Why It Matters Now

The AI industry is running into the limits of brute force. High-quality training data is finite, compute is expensive and scarce, and the cost of each new frontier model keeps climbing. A lab that can get more capability out of less data attacks all three constraints at once.

"A model that's a million times more data efficient is probably a million times easier to put into the economy," Asher Spector told TechCrunch. The logic is that data efficiency is not just a research milestone but a commercial one. Models that can learn from a company's own limited data could be deployed in fields where internet-scale datasets simply do not exist, from specialized manufacturing to rare-disease research.

The round also shows how far investor appetite for frontier AI has stretched. A billion-dollar Series A for a lab with no disclosed product would have been unthinkable a few years ago. Nvidia's participation is a reminder that chipmakers are backing labs across the spectrum, including ones promising to use their hardware more efficiently.

It also reflects a shift in who gets funded. The founders are young researchers rather than veterans spun out of an established lab, and investors are paying a premium for a contrarian technical thesis instead of a proven playbook.

For now, Flapping Airplanes is a research bet with an all-star cap table. The pressure will be to show results that justify a $6 billion price tag before it comes back for more capital.

Source

TechCrunch — "Year-old AI lab Flapping Airplanes raises $1B at a $6B valuation"