Fal, a San Francisco-based startup specializing in AI model inference for video, image, and audio applications, has raised $1.3 billion in its latest funding round, led by Bessemer Venture Partners. This Series F round, combined with an earlier Series E, brought the total raised to $1.65 billion, boosting Fal’s valuation to $16 billion—double its $8 billion valuation just seven months prior in April. Participation in the funding also included Nvidia, Sound Ventures, and T. Rowe Price, while the earlier round had Sequoia as a lead investor.

The surge in valuation reflects the growing demand for inference services, which involve executing pre-trained AI models to generate outputs. Fal has notably benefited from customers leveraging open-source models, contributing to the startup surpassing $800 million in annualized revenue earlier this year. The rapid growth in this sector parallels other inference service providers like Baseten, Fireworks AI, and Modal, which have also secured substantial funding amid high market interest.

Despite the impressive growth, the gross margins for companies providing AI inference remain narrow. This financial pressure is largely due to the substantial expenses involved in acquiring and operating the necessary compute infrastructure to handle AI workloads efficiently. Industry insiders note that the cost challenges are a persistent factor for inference businesses even as demand accelerates.

Fal was co-founded in 2021 by Burkay Gur, formerly of Coinbase and Oracle, and Gorkem Yurtseven, an ex-developer at Amazon. The company’s significant capital influx and rising valuation underscore the strategic importance of AI inference within the rapidly evolving AI ecosystem, positioning Fal as a key player in enabling real-time, scalable deployment of AI applications across various industries.