B200 pricing on GMI Cloud
Every GMI Cloud B200 offering we track, normalized to a per-GPU hourly rate and ranked against 37 providers with current B200 pricing.
The cheapest B200 we currently track is $3.20/hr on Vultr.
GMI Cloud B200 rates
Hourly prices per GPU. Multi-GPU rows show the per-GPU rate, not the instance total.
| Price / GPU-hr | Type | GPUs |
|---|---|---|
| $5.00 | On-demand | 1 |
B200 specifications
- VRAM
- 180GB
- Architecture
- Blackwell
- Memory Bandwidth
- 8000 GB/s
- TDP
- 1000W
B200 on other providers
- Vultr$3.20/hr
- Avian Compute$3.42/hr
- Packet AI$3.75/hr
- Civo$3.79/hr
- Aircloud$4.00/hr
- Together AI$4.09/hr
- UpCloud$4.11/hr
- CoreWeave$4.26/hr
- Verda$4.43/hr
- Deep Infra$4.89/hr
- Google Cloud$4.95/hr
- Beyond.pl$5.10/hr
- Hyperstack$5.10/hr
- Spheron$5.25/hr
- Massed Compute$5.43/hr
- GPU Outlet$5.70/hr
- Lium.io$5.89/hr
- Runpod$5.98/hr
- Hyperbolic$5.99/hr
- Modal$6.25/hr
- OpenRelay$6.30/hr
- Lambda Labs$6.69/hr
- Impossible Cloud$7.06/hr
- Hexgrid Cloud$7.50/hr
- Vast.ai$7.50/hr
- Runcrate$7.67/hr
- Fluence Cloud$7.92/hr
- fal.ai$7.99/hr
- gpu.ai$7.99/hr
- TheAI Cloud$8.40/hr
- Nebius$8.50/hr
- Aquanode$9.35/hr
- Baseten$9.98/hr
- VoltageGPU$11.80/hr
- Oracle Cloud$14.00/hr
- Amazon AWS$14.24/hr
About GMI Cloud
GMI Cloud is a neocloud combining dedicated NVIDIA GPU compute (H100, H200, B200, GB200, with GB300 on pre-order) and a serverless, OpenAI-compatible inference API for text, image, video, and audio models.
Frequently Asked Questions
How much does the B200 cost on GMI Cloud?
GMI Cloud lists the B200 from $5.00 per GPU-hour across 1 tracked configuration. Prices are collected daily — see the table above for the current rates.
Is GMI Cloud the cheapest place to rent the B200?
GMI Cloud ranks #12 of 37 providers we track with current B200 pricing. The cheapest right now is Vultr. Price is only one factor — region availability, network, and storage costs differ between providers.
Which B200 configurations does GMI Cloud offer?
We currently track 1 B200 offering from GMI Cloud, covering the GPU counts, pricing types, and regions listed in the table above. Every price is normalized to a per-GPU hourly rate so it can be compared across providers.