DeepInfra
AI Inference Cloud for Developers
About DeepInfra
DeepInfra is a Palo Alto-based AI infrastructure company focused entirely on inference rather than model training. Founded in 2022 by Nikola Borisov, Yessenzhar Kanapin, and Georgios Papoutsis, the company operates its own GPU infrastructure across eight U.S. data centers and provides OpenAI-compatible APIs for running large language models at scale. As of 2026, DeepInfra supports more than 190 open-source AI models and processes nearly 5 trillion tokens every week for enterprise and developer workloads.
The company gained significant traction as demand shifted from AI training to production inference. In May 2026, DeepInfra raised a $107 million Series B round co-led by 500 Global and Georges Harik, bringing total funding to approximately $153.6M. Since its Series A, DeepInfra reported 25x growth in token volume and has positioned itself as a low-cost, high-throughput alternative to traditional cloud providers for AI deployment.
Founders
Funding History
Frequently Asked Questions
What does DeepInfra do?
DeepInfra provides AI inference infrastructure and APIs that allow developers and enterprises to run large language models and generative AI workloads at scale.
Who founded DeepInfra?
DeepInfra was founded in 2022 by Nikola Borisov, Yessenzhar Kanapin, and Georgios Papoutsis.
How much funding has DeepInfra raised?
DeepInfra has raised approximately $153.6 million across its Pre-Seed, Seed, Funding Round, Series A, and Series B financings.
Who are DeepInfra's investors?
Investors include 500 Global, NVIDIA, Samsung Next, Supermicro, Felicis, A.Capital Ventures, Georges Harik, Crescent Cove, Peak6, and Upper90.
What is DeepInfra known for?
DeepInfra is known for providing low-cost, high-performance AI inference infrastructure for running foundation models in production.
