Artificial intelligence · News
Flapping Airplanes raises $1 billion at $6 billion valuation for AI research
Published:
In brief
San Francisco-based AI research lab Flapping Airplanes has raised $1 billion in a Series A round, lifting its valuation to $6 billion, about four times the $1.5 billion mark it had after a $180 million seed round in January 2026. The year-old startup has no commercial product yet and is focused on making AI model training less dependent on huge amounts of data and computing power.
Flapping Airplanes, an AI research startup based in San Francisco, announced on October 21, 2026, that it had closed a $1 billion Series A funding round at a $6 billion valuation, according to TechCrunch and SiliconANGLE. The round was co-led by venture firms Index Ventures and Lux Capital.
Other participants in the round included returning seed backers Sequoia Capital and GV (formerly Google Ventures), along with Kleiner Perkins, Valor Equity Partners, Atreides Management, Jump Trading, XTX Ventures, and chipmaker Nvidia, SiliconANGLE reported. TechCrunch's list of participants also named A*.
From $1.5 billion to $6 billion in nine months
Flapping Airplanes launched with a $180 million seed round in January 2026 that valued the company at $1.5 billion, according to TechCrunch and TheNextWeb. The new $6 billion valuation represents roughly a fourfold increase in under a year, with SiliconANGLE putting the company's total funding raised to date at $1.18 billion.
TechCrunch described the gap between the seed and Series A rounds as nine months, while SiliconANGLE and TheNextWeb characterized the company itself as roughly one year old. Neither TechCrunch nor SiliconANGLE stated an exact founding date; SiliconANGLE's framing implies the company was founded around a year before the October 2026 announcement.
Who founded the company
Flapping Airplanes was founded by Ben Spector, Asher Spector, and Aidan Smith, according to TechCrunch and SiliconANGLE. Asher Spector holds a statistics Ph.D. from Stanford. Aidan Smith is described as a former Thiel Fellow who previously worked at Neuralink.
Both TechCrunch and SiliconANGLE describe Ben Spector as a computer science Ph.D. student at Stanford who founded the student accelerator Prod while at MIT. SiliconANGLE identifies him as the company's chief executive and reported that investor Index Ventures described the team as roughly two dozen people working in person in San Francisco.
What the company says it is building
Flapping Airplanes describes itself as building what SiliconANGLE called "the world's most complete megakernel framework" for Nvidia graphics processing units. According to TheNextWeb and TechCrunch, the framework is designed to run AI model training and inference simultaneously on the same chips, reducing idle computing time and aiming to let models learn from far less data than current large systems require.
A model that's a million times more data efficient is probably a million times easier to put into the economy.
According to TheNextWeb, Asher Spector also described the company's broader rationale as a belief that demand for AI will outpace available chip and data supply, adding that extracting even modest additional performance from existing chips — around 10% more — would carry significant economic value.
A wave of research-first "AI neolabs"
TechCrunch and TheNextWeb both place Flapping Airplanes within a category of startups sometimes called "AI neolabs": companies that start by focusing on fundamental research rather than shipping a product or generating revenue immediately. TheNextWeb cited two comparable firms, AMI Labs, which it said raised about $1 billion in March 2026, and Mirendil, which raised a $200 million seed round in June 2026.
TechCrunch noted that Flapping Airplanes currently has no product and no revenue, meaning the $6 billion valuation rests on the lab's research progress rather than on sales or a launched product.
| Round | Date | Amount | Valuation |
|---|---|---|---|
| Seed | January 2026 | $180 million | $1.5 billion |
| Series A | October 2026 | $1 billion | $6 billion |
What is not yet known
The company has not disclosed a product launch date, specific performance benchmarks for its megakernel framework, or detailed revenue plans, based on the available reporting. Only SiliconANGLE provided a headcount figure, and the exact founding date was not stated by any of the three outlets reviewed.
Questions
- How much did Flapping Airplanes raise and at what valuation?
- Flapping Airplanes raised $1 billion in a Series A round at a $6 billion valuation, announced October 21, 2026, according to TechCrunch and SiliconANGLE.
- Who led the funding round?
- Index Ventures and Lux Capital co-led the round, with participation from Sequoia Capital, GV, Kleiner Perkins, Valor Equity Partners, Atreides Management, Jump Trading, XTX Ventures, and Nvidia, per SiliconANGLE.
- Who founded Flapping Airplanes?
- The company was founded by Ben Spector, Asher Spector, and Aidan Smith, according to TechCrunch and SiliconANGLE. Asher Spector has a statistics Ph.D. from Stanford, and Aidan Smith is a former Thiel Fellow and ex-Neuralink employee.
- What does Flapping Airplanes actually build?
- It is developing a GPU "megakernel" framework meant to run AI training and inference together, aiming to cut idle computing time and reduce how much data is needed to train AI models, according to SiliconANGLE and TheNextWeb.
- Does Flapping Airplanes have a product on the market?
- No. TechCrunch reported that as of the funding announcement, the company has no commercial product and no revenue.
Sources
This piece was prepared with artificial intelligence and passed an automated source check before publication. If you see an error, ask for a correction. Request a correction or reply · How we make this
Related
Artificial intelligenceArticle
How Generative AI Works: Text, Image, and Code Generation Explained
Generative AI tools such as chatbots, image generators, and coding assistants all convert an input into numerical data, then build an output step by step using patterns learned from huge training datasets. Text and code tools predict the next token in a sequence using transformer models, while image tools like Imagen gradually remove noise from a random pattern until a picture matching a text prompt emerges.
Artificial intelligenceArticle
How AI chatbots like ChatGPT and Claude actually work, explained
AI chatbots are built on large language models (LLMs), neural networks trained to predict the next "token" in a sequence of text. They turn words into numbers, use an attention mechanism to weigh how words relate to each other, and are then fine-tuned with human feedback to behave like helpful assistants rather than raw text predictors.
Artificial intelligenceArticle
How AI Recommendation Algorithms Decide What You See and Buy Online
AI recommendation systems personalize feeds and product suggestions by analyzing what users click, watch, rate and buy, then matching that behavior against patterns from similar users. Platforms such as Netflix, YouTube and Amazon say they use engagement and contextual signals — not, in some cases, demographic data — while regulators in the US and EU have raised concerns about transparency and data collection behind these systems.
Artificial intelligenceArticle
Why AI Language Models Hallucinate: The Research Behind False Answers
AI "hallucinations" happen because language models are trained to predict the next plausible word, not to verify truth, and because the tests used to grade them reward confident guessing over admitting uncertainty. OpenAI researchers argue the problem is partly a mathematical consequence of how models are trained and partly a result of how the AI industry scores them.