GPU Cloud Pricing
Live rates across 5+ providers. No lock-in.
Live GPU cloud pricing for NVIDIA H100, H200, B200, B300, A100, GH200, L40S, RTX 5090, RTX 4090, and RTX PRO 6000 at a fraction of hyperscaler costs. Per-minute billing, no commitments, and instant deployment from certified data centers worldwide. Scale from a single GPU to multi-node clusters on demand. Looking for a per-GPU rental page with specs and use cases? Browse the full GPU rental catalog.
Per-GPU Hourly Rates
Cheapest live on-demand rate per GPU on the Spheron marketplace, with the spot rate below it where one is available. Billed per minute. No commitments.
Need More Than What's Listed?
Reserved Capacity
Commit to a duration, lock in availability and better rates
Custom Clusters
8 to 512+ GPUs, specific hardware, InfiniBand configs on request
Supplier Matchmaking
Spheron sources from its certified data center network, negotiates pricing, handles setup
Tell us your GPU needs and we'll match you with the right provider from our certified data center network.
Typical turnaround: 24–48 hours
Pricing FAQ
The hourly rate covers everything you need to get started: a fully provisioned VM or bare metal instance with your selected GPU, high-speed NVMe SSD storage, network bandwidth, and a dedicated IP address with full root access. There are no hidden fees or surcharges.
No, there is no minimum rental period. Spheron charges with per-minute billing granularity, so you only pay for the exact time you use. Spin up a GPU for a quick test run or keep it running for weeks — it’s entirely up to you.
Yes, we offer competitive volume and enterprise pricing for teams that need multiple GPUs or long-term capacity. Contact our sales team to discuss your requirements and get a custom quote tailored to your workload.
Contact salesWe accept credit cards for traditional payments. We also support stables payments like USDT and USDC, making it easy to pay however you prefer.
Absolutely. You can spin down one GPU instance and spin up another model at any time. There are no lock-in periods or penalties for switching. This flexibility lets you use a cost-effective GPU for development and scale up to a more powerful one for production training.
Spheron’s GPU pricing is significantly cheaper than major cloud providers. For example, an H100 on Spheron starts at $2.65/hr on-demand compared to $6.98/hr on Azure and $7.00/hr on AWS (as of Q1 2026), and spot runs cheaper still. Check our GPU-specific pages for detailed comparison tables against every major provider.
We did not want to spend engineering time comparison-shopping GPU rentals. Spheron found us H200 capacity at the best price in the market, from a Tier 3+ partner they had already vetted, with long-term commitment terms negotiated on our behalf and the rental validated before we went live. When we need more capacity, or anything operational with the data center, we message their team directly. That lets us stay focused on the product.
Jaipal Singh, Chief Technology Officer, Prem AIMore GPU Pricing Guides
More GPU Pricing guides →GPU Cloud Pricing Comparison 2026
H100 from $2.01/hr, B200 spot at $5.34/hr: full price table across 15+ providers.
AWS H100 Pricing 2026
P5 H100 on-demand breakdown vs Spheron H100 from $2.64/hr on-demand.
NVIDIA B200 Cloud Pricing 2026
On-demand, spot, and reserved B200 rates compared across every major provider.
Rent From $0.53/hr
That is the cheapest on-demand rate live on the marketplace right now. Spot runs cheaper but can be reclaimed at any time. Either way you are billed per minute, with no minimum rental period and no contracts.