Skip to content
Neural Daily

BusinessNews

Flapping Airplanes raises $1B to make AI far more efficient

🇭🇺 Olvasd el magyarul →

Year-old Flapping Airplanes raised a $1 billion Series A at a $6 billion valuation, co-led by Index Ventures and Lux Capital. The lab aims to train large models with far less data and compute.

By: Mikola Máté Patrik · · 2 min read

The three Flapping Airplanes co-founders posing in a white photo studio, one seated on the floor, one in a white armchair and one standing
Image: Index Ventures · Press kit · Founder photo from the Index Ventures announcement post · Source

What happened

San Francisco-based Flapping Airplanes announced a $1 billion Series A on October 21, co-led by Index Ventures and Lux Capital (Index Ventures). The round values the company at $6 billion, four times its $1.5 billion seed valuation from January. Earlier backers Sequoia, GV and Kleiner Perkins also took part (TechCrunch).

The lab was founded by Aidan Smith and brothers Asher and Ben Spector. The team is a couple dozen people, all working in person in San Francisco. They hire for trajectory and skill over experience: per Index, one engineer joined while still in high school, and another had never trained a model before leading the company's entire pre-training run.

What they are building

Flapping Airplanes believes AI can be orders of magnitude more efficient. The team works across the stack, from GPU kernels through pre- and post-training to inference. Its core tool is a megakernel framework that runs an entire training or inference workload as a single GPU program, removing idle time between steps.

The founders say the framework can overlap memory-bound inference with compute-bound training, so the chip rarely sits idle. The goal is models that learn without ingesting half the internet. As the founders put it, a model that is a million times more data efficient is probably a million times easier to put into the economy.

Why it matters

While the big labs raise money for ever larger data centers, this round is a bet that the next leap comes from efficiency. Index points to markets where AI is still too expensive or data is scarce: robotics, trading and life sciences. The company has not yet shown a public model or benchmark, so for now the $6 billion is a bet on the team and the approach.

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles