The GPU cloud provider secured $223M in equity led by ARCHIV and a $445M credit facility led by Taiwanese bank CTBC to grow its hardware footprint in the U.S. and Asia. Nvidia participated in the equity round and partners with GMI Cloud through an AI compute program; GMI is one of seven providers building facilities entirely on the hardware and software stack recommended by Nvidia.

The company reports that its contracted annualized revenue has reached more than $600M, an increase of over 9x since the end of 2025. GMI Cloud currently processes 1 trillion AI inference tokens per day for customers including OpenRouter, Fireworks AI and Higgsfield.

Sign in to suggest edits

Key sources

  1. SOURCEmarketbrief.now
  2. SOURCE@wallstengine“processing 1 trillion AI inference tokens per day, up 20x in six months”x.com
  3. SOURCEhuggingnewshuggingnews.com
Markdown