Fal raises $1.3B Series F led by Bessemer, doubling its valuation to $16B

Bessemer, an existing investor that has backed Fal in every round since its Series B two years ago, led the new round[2]. NVIDIA, Sound Ventures, HOF Capital and T. Rowe Price also took part[4]. The round closed in October, according to The Next Web[3].

Fal reportedly reached $800 million in annualised revenue earlier this year[3]. Chief executive Burkay Gur said that in the summer "we were not able to serve our customers and had to rate limit them"[3]. Fal says the new money pays for inference capacity and the team to build what comes next, funding work across its ecosystem in 2027[1]. It currently has about 185 staff[3].

Bessemer argues that serving image and video models is not the same job as serving chatbots: language models produce one token at a time, while image and video models refine a whole frame over many passes, which changes how memory, compute and latency have to be handled[2]. It says Adobe, Canva, Shopify, World Labs and Amazon have moved production volume to Fal, and that Fal has grown from an inference engine into workflow tools, an API to hundreds of models, custom infrastructure for enterprises' own weights, and post-training models with the labs that make them[2].

A $16 billion price for a company that mostly runs other people's models shows where investors think generative media's money will settle: in the serving layer that studios, ad agencies and e-commerce brands plug into their production pipelines[1]. Fal's summer rate limits also show that compute, not customers, is the constraint for now.