Flapping Airplanes raises $1 billion at $6 billion valuation in AI research funding round
A new wave of investor money is continuing to flow into artificial intelligence research startups with limited commercial track records. Flapping Airplanes, founded in late 2025, has secured $1 billion at a $6 billion valuation as venture firms back efforts to make AI models and chips more efficient.
Highlights
- Flapping Airplanes raised $1 billion at a $6 billion valuation in a funding round led by Index Ventures and Lux Capital, as reported by Bloomberg.
- The new valuation has quadrupled since January, with backing from Sequoia Capital, Alphabet Inc.'s GV, and Kleiner Perkins, reflecting heightened investor interest in AI neolabs.
- Co-founder Asher Spector stated the company targets improved AI model training efficiency and chip performance, addressing infrastructure cost concerns amid anticipated chip and data supply constraints.
Funding round expands backing for AI neolabs
As first reported by Bloomberg, the financing is set to be announced Wednesday and is led by Index Ventures and Lux Capital. Other investors in the round include Sequoia Capital, Alphabet Inc.'s GV and Kleiner Perkins, and the valuation, including the new capital, has quadrupled from a prior round in January.Flapping Airplanes is part of a growing group of AI "neolabs", startups that focus first on research instead of quickly launching products and building revenue. Investors have poured billions of dollars into companies pursuing AI-led scientific advances, self-improving systems and related technologies.
Efficiency pitch targets AI infrastructure costs
Co-founder Asher Spector says the company is developing methods to train AI models using less data and less computing power, while also improving the efficiency of chips used to build and run those models. The strategy comes as spending on chips and data centers for AI development faces greater scrutiny.Spector says the company's core view is that demand for AI is going to outpace chip and data supply. He says a company that can extract more performance from existing inputs could create significant economic value, adding that even a 10% performance gain from each chip would be tremendously valuable given current infrastructure spending.
Our earlier article on the mounting pressures facing U.S. AI data centers explained that many projects are being delayed or reconsidered as financing costs rise and construction timelines slip. We also noted that power constraints and local opposition are increasingly limiting how fast computing capacity can expand, testing investor confidence in the broader AI infrastructure buildout.