Source: Multimodal AI startup Fal.ai has already raised a valuation of over $4 billion

Fal.ai, a startup that hosts image, video and audio AI models for developers, has closed on a new round of funding that values ​​the company at more than $4 billion, four people familiar with the deal said. The company has raised about $250 million, according to two people familiar with the matter.

Kleiner Perkins and Sequoia are the lead investors in the round, sources said. Farr did not respond to requests for comment. Sequoia and Kleiner Perkins declined to comment.

The new round comes less than three months after Fal announced a $125 million Series C led by Meritech at a valuation of $1.5 billion. At the time, the company had more than $95 million in revenue and its platform was used by more than 2 million developers, Todd Jackson, a partner at First Round Capital, wrote on LinkedIn. This was a significant growth from a year ago, when TechCrunch reported that Fal had $10 million in annual recurring revenue (ARR) and 500,000 developers.

Because Fal provides the infrastructure layer for multimodal AI models (as well as media-specific AI models), the company’s explosive growth is directly tied to user adoption of the applications built on top of it. And multimodal AI, especially video, is in high demand right now, as evidenced by the meteoric rise in popularity of OpenAI’s Sora, which rose to the top of the US App Store even faster than ChatGPT. This strong consumer demand for applications like Sora highlights the market potential of Fal’s products.

Fal offers developers over 600 images, videos, audio, and 3D models, and boasts thousands of Nvidia H100 and H200 GPUs in its cloud, fine-tuned for fast inference. It also provides tools for customizing the model. Its products include access via APIs, flexible serverless products, or enterprise-ready compute clusters. There are certainly other competitors offering model and app hosting services (Microsoft, Google,
(CoreWeave, to name a few), VCs like Jackson say Fal’s singular focus on media and multimodal is a competitive selling point.

The startup’s customers range from individual developers to large enterprises including Adobe, Canva, Perplexity, and Shopify. Common use cases include media creation for advertising, e-commerce, and gaming content.

The startup was co-founded in 2021 by Burkay Gur, a former Coinbase machine learning leader and Oracle engineer, and Gorkem Yurtseven, a former Amazon developer. Gur and Yurtseven saw an opportunity for personalized multimedia generation. While other engineers pursued LLM, they focused on optimizing stable diffusion for speed and scale, and subsequently expanded to hosting many other similar models.

Farr has raised nearly $200 million to date, according to PitchBook data. The company’s existing investors include Bessemer Venture Partners, Kindred Ventures, Andreessen Horowitz, Notable Capital, First Round Capital, Unusual Ventures and Village Global.