AI cloud provider DeepInfra secured $107 million in a Series B funding round, funds it will use to expand its inference cloud platform and global capacity.

The funding round was co-led by 500 Global and Georges Harik, and saw participation from A.Capital Ventures, Crescent Cove, Felicis, Nvidia, Peak6, Samsung Next, Supermicro, and Upper90.

"When we started this company nearly four years ago, we believed inference would become the dominant driver of AI workloads," DeepInfra CEO Nikola Borisov wrote on LinkedIn. "Today, we’re at that inflection point. Open-source models are rapidly reaching parity with proprietary systems, and agent-based systems are driving continuous, high-volume demand. Inference is no longer a thin layer – it’s the system constraint. Most cloud platforms weren’t built for this. So we built our platform from the ground up for high-throughput inference, optimizing for cost, performance, and security at production scale."

DeepInfra currently operates out of eight U.S. data centers, but is rolling out more locations globally. Details about its current locations or planned new locations have not been shared.

The company already offers Nvidia Blackwell graphic processing units (GPUs), and will be adding the upcoming Vera Rubin GPUs in due course.

Founded in 2022, DeepInfra is based in Palo Alto, California. The company completed an $18 million Series A funding round in April 2025.