Provider pricing

GMI Cloud AI API pricing

GMI Cloud is an NVIDIA Preferred Partner GPU cloud founded in 2023 and headquartered in San Jose, California. Offers H100, H200, and GB200 capacity with bare-metal and Kubernetes-managed options.

Founded2023
HeadquartersSan Jose, CA
Tracked modelsDirectory

Best for

Teams scaling onto next-gen Blackwell GPUs early.

Strengths

  • NVIDIA Preferred Partner with priority B200/GB200 supply
  • Kubernetes-managed inference with auto-scaling
  • Reserved H100 clusters from $2.10/hr
  • Listed in the TokenCost AI providers directory