GMI Cloud has raised $668 million in new financing, combining a $223 million Series B equity round with a $445 million credit facility, as the AI infrastructure company expands GPU capacity across the United States, Taiwan, and the broader Asia-Pacific region. The Series B was led by ARCHIV, a San Francisco-based investment firm focused on AI and robotics, with participation from NVIDIA. Other investors from the Asia-Pacific region included DSC Investment, Trend Micro, KB Investment, Kyobo Life, and KT Corporation.
The $445 million credit facility was led by CTBC, providing GMI Cloud with additional capital to expand its infrastructure footprint as demand for GPU computing and AI inference services continues to increase.
The funding follows a period of substantial commercial growth for the Mountain View, California-based company. GMI Cloud said its contracted annual recurring revenue has surpassed $600 million and reached more than nine times its level at the end of 2025.
The company’s live ARR from infrastructure currently in production has increased more than 4.5 times over the same period, while its inference platform is now processing approximately 4 trillion tokens per week.
GMI Cloud plans to use the financing to expand GPU capacity across the U.S., Taiwan, and other Asia-Pacific markets, while continuing to develop its inference services and adding employees as the business scales.
The expansion builds on GMI Cloud’s Taiwan AI Factory, which was announced in 2025 as the company’s first major infrastructure facility in Asia.
GMI Cloud has also announced a sovereign AI initiative in Japan, expanding its strategy of building infrastructure closer to customers that may have data-residency, latency, security, or regulatory requirements.
The company is positioning its infrastructure platform around the idea that AI companies increasingly need compute capacity spanning both the U.S. and Asia rather than operating exclusively in one region.
U.S.-based AI developers and hyperscalers may require infrastructure closer to customers and operations in Asia, while enterprises throughout Asia-Pacific increasingly want production AI workloads hosted locally under domestic compliance requirements.
GMI Cloud operates its infrastructure through a unified platform that allows customers to deploy workloads based on the locations of their users, data, and regulatory requirements.
The company also sees its proximity to Taiwan’s technology manufacturing ecosystem as an important part of its infrastructure strategy.
Taiwan is a critical center for global AI server and semiconductor manufacturing, and GMI Cloud said its relationships within that supply chain provide a more predictable path between ordering hardware and deploying functioning GPU clusters.
That supply-chain position is becoming increasingly important as AI infrastructure providers compete not only on GPU availability but also on how quickly new systems can be delivered and placed into production.
GMI Cloud specifically highlighted its ability to deploy advanced NVIDIA GB200 and GB300 NVL72 systems, with AI infrastructure company Fireworks citing GMI Cloud as one of its more reliable providers for those platforms.
GMI Cloud’s notable customers also include Higgsfield, Nous Research, OpenRouter, Reflection, Cartesia, Trend Micro, and Utopai Studios, alongside Fireworks.
The company’s platform extends beyond raw GPU infrastructure.
GMI Cloud describes itself as an AI-native cloud built around compute, inference, and AI agents, combining GPU clusters, optimized inference services, and agent infrastructure on a unified cloud environment.
That strategy puts GMI Cloud in a growing group of infrastructure providers seeking to compete beyond simply renting GPUs by offering software and services that help companies deploy and operate production AI workloads.
GMI Cloud was founded in 2021 and is headquartered in Mountain View. The company currently operates GPU infrastructure across the United States and Asia-Pacific.
The new financing gives GMI Cloud significantly more capital to build out that footprint at a time when AI model developers, application companies, and enterprises are competing for access to advanced computing capacity.
With more than $600 million in contracted ARR, GMI Cloud is also moving from an emerging infrastructure provider toward operating AI infrastructure at significantly greater scale.
KEY QUOTES:
“Our customers are scaling faster than ever, and they need infrastructure that keeps pace. AI is driving a new renaissance, and reliable compute is its foundation. Our goal is to build that foundation across continents, with an ecosystem of products on top of it.
In AI infrastructure, a delivery date is a promise. Customers plan launches, hiring, and revenue around it. Our place in Taiwan’s supply chain is how we keep that promise, cluster after cluster globally.”
Alex Yeh, founder and CEO of GMI Cloud
“Capacity that arrives late is capacity we can’t use. GMI Cloud has been one of our strongest and most reliable providers across NVIDIA GB200 and GB300 NVL72 systems.”
Chenyu Zhao, co-founder of Fireworks
