GMI Cloud Secures 8x ARR Growth as Global AI Infrastructure Footprint Expands

Following 2.4x growth in live ARR in the first half, GMI Cloud has secured 8x contracted ARR for year-end as surging AI demand fuels its global expansion

MOUNTAIN VIEW, Calif., July 21, 2026 /PRNewswire/ -- GMI Cloud, a leading AI-native cloud provider delivering high-performance GPU infrastructure and inference services, today announced record growth in 2026, underscoring surging global demand for production-scale AI compute. Live ARR, revenue from GPU capacity running in production today, grew by 2.4x in the first half of 2026 over year-end 2025. On top of that, contracted ARR as of today from signed customer commitments, including multiple multi-year agreements with leading AI companies, has surpassed $500 million. This represents more than 8x GMI Cloud's year-end 2025 ARR and will  progressively convert to live ARR as additional power capacity comes online this year.

This commercial momentum is accelerating GMI Cloud's transition into a scaled, global AI infrastructure platform. The company has expanded its data center footprint from the United States into Taiwan and Thailand, with further expansion underway across Asia-Pacific, supported by partnerships with Macnica, Compal, and Taiwan Mobile. As part of this global strategy, GMI Cloud announced a sovereign AI infrastructure initiative in Japan, anchored by a 200 MW first-phase deployment in 2028.

"GMI Cloud has built its platform for exactly this moment," said Alex Yeh, CEO of GMI Cloud. "Our growth shows that customers need more than access to GPUs, they need an AI-native cloud that can turn infrastructure into production outcomes. We are scaling our capacity, inference services, and global footprint to meet that demand."

This expanding footprint is already supporting customers across frontier AI, sovereign AI, cinematic production, legal technology, cybersecurity, agentic AI, and generative AI. GMI Cloud's customers and partners include Fireworks, Higgsfield, Trend Micro, Wistron, and Utopai Studios.

"As enterprises move beyond renting general-purpose AI and build specialized intelligence on their own data, reliable infrastructure becomes mission critical," said Dmytro Dzhulgako, Co-founder at Fireworks. "GMI Cloud's high-performance compute helps us deliver production-scale inference with the speed, reliability, and flexibility our customers need to build AI that's uniquely their own."

GMI Cloud's growth extends beyond GPU infrastructure. Its inference business grew 34x in four weeks and now processes approximately 2.5 trillion tokens each week, reflecting accelerating demand for high-performance, production-scale inference.

GMI Cloud is also a top-three token provider on OpenRouter, serving AI customers seeking U.S. and APAC token endpoints. Its inference platform supports leading AI laboratories, agentic AI companies, and generative AI brands that require scalable capacity for production workloads.

By combining NVIDIA GPU infrastructure with AI-optimized cloud and inference services, GMI Cloud enables customers to scale model training, post-training, fine-tuning, and high-volume agentic inference while supporting secure, regionally compliant sovereign AI deployments.

GMI Cloud is set to continue this trajectory as it expands its infrastructure presence from the United States across Asia and supports customers deploying production-scale AI workloads across U.S. and APAC markets. For more information, visit www.gmicloud.ai.

About GMI Cloud

GMI Cloud is an AI-native cloud infrastructure company powering the next generation of AI applications. The company provides high-performance GPU infrastructure, Model-as-a-Service, dedicated endpoints, and AI workload deployment solutions for developers and enterprises building production AI systems. GMI Cloud helps teams move from experimentation to production with scalable compute, flexible infrastructure, and an ecosystem built for modern AI builders. For more information, visit gmicloud.ai.

Media Contact:
Angela Baldwin
Baldwin PR on behalf of GMI Cloud
[email protected]

SOURCE GMI Cloud