Quick Facts
- DeepInfra raised $107M Series B led by 500 Global and Georges Harik, with participation from Nvidia, Samsung Next, and others
- Company processes nearly five trillion tokens per week, growing volume 25x since Series A round
- Revenue tripled in 2026 while supporting 190+ open-source models across eight U.S. data centers
DeepInfra closed a $107 million Series B funding round to expand its dedicated inference cloud for open-source AI models. The round was led by 500 Global and Georges Harik, one of Google’s first cloud engineers.
Major investors including Nvidia Corp., Samsung Next, Supermicro Computer Inc., A.Capital Ventures, Crescent Cove, Felicis, Peak6 and Upper90 participated in the round. The funding brings DeepInfra’s total raised to over $133 million since its 2022 founding.
The company now processes nearly five trillion tokens per week, representing 25x growth since its Series A. Revenue has tripled since the beginning of 2026, while the company scaled processing volume more than 8,000 times since its seed round.
“When we launched nearly four years ago, we believed inference would become the dominant driver of enterprise AI workloads – and we are now at this inflection point,” said CEO Nikola Borisov. “Open-source models are rapidly reaching parity with proprietary systems, unlocking a new wave of innovation at a fraction of the cost.”
DeepInfra supports 190+ open-source models through APIs priced to undercut OpenAI and Anthropic by five to ten times. The company was advertising $1 per million token rates in 2023 against GPT-4 Turbo’s $10 rate. Nearly 30% of weekly token volume comes from AI agents like OpenClaw.
The startup owns GPU infrastructure across eight U.S. data centers with global expansion planned. DeepInfra works with Nvidia to optimize performance, including early deployment of Nvidia Dynamo distributed-inference software that unlocks 20x improvements in inference cost efficiency.
“The demand for AI inference is going through the roof, and engineers and developers are finding that they need faster, more flexible and reliable infrastructure to support it,” said Tony Wang, Managing Partner at 500 Global. “Purpose-built inference will be fundamental to the next phase of AI.”
The funding will accelerate three priorities: expanding global compute capacity, deepening developer tooling, and supporting the next generation of open-source and agentic models. Inference workloads will account for roughly two-thirds of all compute in 2026, up from a third in 2023.
DeepInfra was founded by Nikola Borisov, Yessenzhar Kanapin, and Georgios Papoutsis, former engineers at imo.im who built systems serving 200 million users. The company offers SOC 2 and ISO 27001 certification with zero data retention policy.
Read more: Deepinfra lands $107M in funding to build out its dedicated inference cloud for open-source models
This article was written by an AI agent. Spotted an error? Send a correction and we will fix it.