Applied Compute Raises $350 Million at $3.25 Billion Valuation

By Grace Kim

Applied Compute raised $350 million at a $3.25 billion valuation, according to FinSMEs. The San Francisco, CA-based AI cloud and infrastructure startup plans to use the proceeds to expand compute capacity, grow its team and accelerate enterprise deployments.

FinSMEs’ April 11, 2026 report did not identify the round’s structure, Applied Compute’s cumulative funding or the dilution attached to the valuation. The figures establish the size of this investment and the company’s valuation, but not the founders’ or investors’ resulting ownership.

What does the $350 million fund?

Elad Gil led the round, according to FinSMEs. The participating firms were Kleiner Perkins, Lux Capital, Hanabi Capital, Nvidia, Blackstone, Shopify Ventures, HubSpot Ventures, Diffusion, Definition, Timeless, Layer Global, Neo, Abstract, Nova, Quantum Light, First Harmonic and Swish.

According to FinSMEs, Applied Compute set three priorities for the money:

• Expand compute capacity.
• Grow the team.
• Accelerate deployment in the enterprise.

FinSMEs did not assign dollar amounts to those priorities or say how much would cover compute, hiring and deployment. The allocation targets both the infrastructure needed to serve customers and the people needed to put their models into production.

How quickly has the business scaled?

Applied Compute has grown to $100 million in annualized revenue run rate less than 18 months after its 2025 founding, FinSMEs reported. Since its previous fundraise, the company has more than doubled the number of enterprises it serves.

The report did not disclose an absolute customer count, a prior customer baseline or a reconciliation of the run rate to recognized revenue. It also gave no revenue by customer, contract value, pricing or retention data. The two disclosed growth markers are a nine-figure annualized revenue pace and an enterprise count that has more than doubled.

Applied Compute was founded in 2025 by Yash Patil, Rhythm Garg and Linden Li, whom FinSMEs identified as former OpenAI technical researchers. The short operating history means the $3.25 billion valuation rests on less than 18 months of activity.

In that period, the startup has combined a $100 million annualized revenue run rate with growth across its enterprise base. Those are the operating markers investors are backing at the new valuation.

What does Applied Compute sell?

FinSMEs described Applied Compute as a developer of custom open-source AI models for enterprises. It also described the startup as an AI cloud and infrastructure company that “focuses on helping enterprises build, train, and own their custom AI models.”

That wording puts Applied Compute in two roles: it develops models tailored to an enterprise’s work, and it provides the cloud and infrastructure layer for building, training, running and owning them. FinSMEs said companies use the offering to build and run AI tailored to their operations.

The combination is central to the company’s pitch: customers can own the models and use Applied Compute’s infrastructure to operate them. The report did not disclose pricing, gross margins, model-training costs or the mix between development and infrastructure work.

In short, Applied Compute is selling control over enterprise models alongside the infrastructure required to run them.

Who is using Applied Compute?

FinSMEs named eight companies using Applied Compute to build and run AI tailored to their work: Microsoft, Shopify, Mercor, Harvey, DoorDash, Cognition, Clay and Decagon.

The adopters span life sciences, cybersecurity, hardware design, legal, financial services, logistics and other industries, according to FinSMEs. The customer list gives the startup exposure across eight named organizations, but the report did not say whether any one company accounts for a material share of revenue.

It also disclosed no contract sizes, renewal rates, deployment scale or sector-level revenue. The fastest route to assessing revenue quality would be data on customer concentration and the size of individual deployments, neither of which appeared in the report.

For Applied Compute, the next proof points are whether a customer base that has more than doubled can sustain the $100 million annualized run rate, and whether the new compute capacity supports faster deployment without changing the enterprise model. The $3.25 billion valuation now requires the San Francisco startup to turn those early operating markers into durable, repeatable revenue.