The cheapest GH200 on the table right now is $2.29 per GPU-hour at Lambda Labs, out of 5 offers from 5 providers. The GH200 carries 96 GB of VRAM on NVIDIA's Hopper generation.
On-demand pricing for the same card spans $2.29 to $6.50 per GPU-hour — a 2.8× spread between the cheapest and the most expensive provider for identical silicon. That gap is the entire reason this table exists.
| Provider | Tier | $/GPU-hr | Pricing | Node | Commit | In stock | |
|---|---|---|---|---|---|---|---|
| Lambda Labs | dedicated | $2.29 | on-demand | 1× GPU | — | — | rent |
| Prime Intellect | dedicated | $2.29 | on-demand | 1× GPU | — | yes | rent |
| Spheron | dedicated | $3.02 | on-demand | 1× GPU | — | — | rent |
| Denvr Dataworks | dedicated | $3.87 | on-demand | 1× GPU | — | — | rent |
| CoreWeave | dedicated | $6.50 | on-demand | 1× GPU | — | — | rent |
How much does a GH200 cost per hour?
As of the latest scrape, $2.29 per GPU-hour at Lambda Labs is the cheapest GH200 offer across 5 tracked cloud providers. The cheapest on-demand rate is $2.29 per GPU-hour at Lambda Labs, and the most expensive on-demand rate is $6.50 at CoreWeave.
Which cloud is cheapest for the GH200?
Lambda Labs at $2.29 per GPU-hour. Prices move constantly, so this page is regenerated from the live scrape every 15 minutes. Note that marketplace tiers are peer hardware and are not directly comparable to dedicated capacity.
How many GB of VRAM does the GH200 have?
96 GB, on the Hopper architecture. Multi-GPU instances multiply that: an 8× GH200 node exposes 768 GB of GPU memory in total.