Fal, the San Francisco startup that runs AI image, video and audio models for other companies, just closed a $1.3 billion Series F that puts the Fal valuation at $16 billion. That is double the $8 billion price tag it carried in April, and it shows how fast money is flowing into AI inference, the work of actually running trained models for paying customers.
Inside the Fal Funding Round
Bessemer Venture Partners led the new round, with Nvidia joining as a strategic investor. Sound Ventures and T. Rowe Price also took part, according to reporting on the deal.
Fal announced its Series F together with an earlier Series E that Sequoia Capital led this year but had not publicized. Across the two rounds, the company raised a combined $1.65 billion.
In its own short announcement, Fal said the money will pay for two things: more inference capacity to run models at scale and a bigger team to build what comes next. The company said the rounds fund its work through 2027 and give it room for new research.
The deal at a glance
| Detail | What was reported |
|---|---|
| Latest round | Series F, $1.3 billion |
| Lead investor | Bessemer Venture Partners |
| Other investors | Nvidia (strategic), Sound Ventures, T. Rowe Price |
| Valuation | $16 billion, up from $8 billion in April |
| Series E and F combined | $1.65 billion (Series E led by Sequoia) |
| Annualized revenue | More than $800 million earlier this year, per Bloomberg |
Why Fal’s Valuation Doubled So Fast
The short answer is demand. Companies that want to use open-source image and video models often do not want to buy and manage their own GPUs, so they pay a service like Fal to host the models and return results quickly.
Fal says entertainment studios, ad agencies, online stores, game developers and marketing teams now run image, video and audio models inside the pipelines that deliver their final work. Some of those models are Fal’s own, which the company has post-trained to be faster without giving up quality.
Bloomberg reported that Fal passed $800 million in annualized revenue earlier this year. A business growing that quickly in a hot market is exactly what late-stage investors are hunting for right now.
A Crowded Race to Run AI Models
Fal is not alone. Other inference providers, including Baseten, Fireworks AI and Modal, have also been growing quickly and raising big rounds at higher valuations.
There is a catch, though. People familiar with these companies’ finances say their gross margins remain thin, because buying or leasing the computing power behind every request is expensive.
That is one reason Nvidia’s spot on the cap table stands out. Inference companies are among the biggest buyers of its chips, so a closer tie can matter on both sides of the deal.
Who started Fal
Burkay Gur and Gorkem Yurtseven co-founded Fal in 2021. Gur previously led machine learning work at Coinbase and was an engineer at Oracle, while Yurtseven was a developer at Amazon.
What This Means for You
If you use AI tools to make images, clips or voiceovers, there is a good chance some of them already run on a service like Fal behind the scenes. More funding usually means more capacity, which can translate into faster results and access to newer models.
Here is what to keep in mind:
- Expect more choice in AI video and image apps, since inference hosts make it cheaper for small teams to ship products built on open models. Our guide to AI image generators explains what is happening under the hood.
- Watch pricing. Thin margins across inference providers could mean price changes or usage limits if computing costs stay high.
- If you build products, compare providers on speed, model selection and cost per request rather than brand alone. Our explainer on inference scaling covers why running models is getting so costly.
- Follow the money. Big rounds like this one, and recent deals such as the Traversal funding round, show where investors think AI spending is headed.
Our takeaway: training models gets the headlines, but running them every day for millions of requests is where a lot of the money is now going.
For creators who want to try generative video, our comparison of video generation models is a good place to start. You can find more funding stories on our AI news page.
Sources
Primary source: Fal’s Series E and F announcement. Additional reporting from TechCrunch, which also cites figures first reported by Bloomberg.
Conclusion
Fal’s $1.3 billion Series F doubles its valuation to $16 billion in roughly seven months and brings its combined new funding to $1.65 billion. With Bessemer leading and Nvidia on board, the startup now has fresh cash to add capacity and staff through 2027. The bigger story is that demand for running AI models is booming, even as thin margins remain a real test for every company in this race.
