Applied Compute in Talks to Double Valuation to $3 Billion on Open-Source Demand

Applied Compute, a year-old startup that helps companies run and customize open-source models with their own data, is discussing a new funding round that would value it at around $3 billion, more than double its valuation from a round announced four months ago, according to a person with knowledge of the funding.

The financing follows strong revenue growth in recent months. Applied Compute is currently generating around $50 million in annualized revenue, the person said, up nearly four times from what its CEO Yash Patil has said it was generating last November.

Investor Elad Gil is in talks to lead the round, which will likely be the hundreds of millions of dollars, the person said. It’s not clear if the valuation includes the new money. The funding round hasn’t yet closed, and its terms could change.

The fundraise also comes amid rising interest from companies looking to cut AI costs by using and customizing open-source models instead of shelling out money for Anthropic and OpenAI models. Such demand has also benefited startups like OpenRouter, which helps developers access and choose between hundreds of AI models, and inference providers such as Baseten and Fireworks, which help application developers train, customize and run primarily open-source AI models.

Former OpenAI researchers Patil, Rhythm Garg and Linden Li founded Applied Compute in 2025. The San Francisco-based startup helps businesses develop and train custom AI models for specific fields, like financial services and law. It does so through a technique called reinforcement learning, which works by rewarding models for accomplishing certain goals and penalizing them for other behaviors.

Applied Compute has said that it works with customers including food delivery service DoorDash, coding startup Cognition and data labeling firm Mercor to use those custom models to develop AI agents that can take actions on behalf of workers at those customers. Applied Compute is also working on methods to help agents learn and improve from real-world experiences, known as continual learning.

Applied Compute makes money through selling consulting services to businesses to help them customize models, as well as charging those businesses for compute to then run those models on the startup’s infrastructure.

The startup has previously raised $160 million in funding from investors including Kleiner Perkins, Benchmark and Sequoia Capital.