AI Inference Startup Fal Doubles Valuation to $16 Billion in $1.3 Billion Round

Bessemer Venture Partners led the round, a Series F. Nvidia, Sound Ventures and T. Rowe Price also invested. Fal announced its Series E and Series F together and said the two rounds raised a combined $1.65 billion. Sequoia led the earlier round.

According to Bloomberg, Fal passed $800 million in annualised revenue earlier this year.

A Fast Climb

The valuation has risen steeply. In July 2025 Fal raised a $125 million Series C at $1.5 billion. In December 2025 it raised $140 million in a Sequoia-led Series D at $4.5 billion, with Nvidia's venture arm and Bessemer taking part. In September 2026 The Information reported that Fal was talking to investors about a valuation of about $15 billion, and said the target could go higher.

Fal was founded in 2021 by Burkay Gur, who previously led machine learning at Coinbase and worked as an engineer at Oracle, and Gorkem Yurtseven, a former Amazon developer.

Why Inference Is Drawing Capital

Inference means running a model that has already been trained so that it produces outputs. It has become one of the busiest parts of the AI market. TechCrunch said demand is especially strong among customers that build on open-source models, which they can run through specialist hosts like Fal instead of building their own infrastructure.

Fal is not the only one raising. Rival inference providers including Baseten, Fireworks AI and Modal are also growing quickly and raising large rounds at higher valuations. The business has a structural weakness, though. TechCrunch, citing several people familiar with these companies' finances, reported that gross margins stay thin because buying or leasing compute is expensive.

Nvidia's investment fits a familiar pattern. The chipmaker has backed many companies that buy its hardware or depend on it, and that has helped tie the fast-growing inference layer to its GPUs.

For enterprise buyers, the round suggests that generative media is now a large category in its own right, and that the providers handling it at scale are well funded. For investors, the main open question is the same one hanging over the wider inference sector: whether revenue growth can turn into durable margins while compute costs stay high.