The cheapest RTX 4090 on the table right now is $0.160 per GPU-hour at Salad Cloud, out of 15 offers from 9 providers. The RTX 4090 carries 24 GB of VRAM on NVIDIA's Ada generation. That is peer/marketplace capacity — rented from other users' machines, and not directly comparable to a dedicated instance.
On-demand pricing for the same card spans $0.160 to $0.907 per GPU-hour — a 5.7× spread between the cheapest and the most expensive provider for identical silicon. That gap is the entire reason this table exists.
Cheapest by pricing model: on-demand $0.160 at Salad Cloud · spot $0.267 at Vast.ai.
| Provider | Tier | $/GPU-hr | Pricing | Node | Commit | In stock | |
|---|---|---|---|---|---|---|---|
| Salad Cloud | marketplace | $0.160 | on-demand | 1× GPU | — | yes | rent |
| Vast.ai | marketplace | $0.267 | spot | 1–4× GPU | — | yes | rent |
| Akash | marketplace | $0.280 | on-demand | 1× GPU | — | yes | rent |
| Novita | dedicated | $0.340 | spot | 1× GPU | — | — | rent |
| Vast.ai | marketplace | $0.366 | on-demand | 1× GPU | — | yes | rent |
| Vast.ai | marketplace | $0.401 | on-demand | 2, 4× GPU | — | yes | rent |
| Octa | marketplace | $0.420 | on-demand | 1× GPU | — | yes | rent |
| Spheron | dedicated | $0.580 | on-demand | 1× GPU | — | — | rent |
| Novita | dedicated | $0.670 | on-demand | 1× GPU | — | — | rent |
| Beam | dedicated | $0.691 | on-demand | 1× GPU | — | yes | rent |
| Runpod | dedicated | $0.890 | on-demand | 1× GPU | — | yes | rent |
| Prime Intellect | dedicated | $0.907 | on-demand | 1× GPU | — | yes | rent |
How much does an RTX 4090 cost per hour?
As of the latest scrape, $0.160 per GPU-hour at Salad Cloud is the cheapest RTX 4090 offer across 9 tracked cloud providers. The cheapest on-demand rate is $0.160 per GPU-hour at Salad Cloud, and the most expensive on-demand rate is $0.907 at Prime Intellect.
Which cloud is cheapest for the RTX 4090?
Salad Cloud at $0.160 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.
Is spot RTX 4090 capacity cheaper?
Yes — the cheapest spot RTX 4090 is $0.267 per GPU-hour at Vast.ai, +67% versus the cheapest on-demand rate of $0.160. Spot instances can be reclaimed by the provider at any time, so they suit checkpointed training and batch inference rather than long-lived services.
How many GB of VRAM does the RTX 4090 have?
24 GB, on the Ada architecture. Multi-GPU instances multiply that: an 8× RTX 4090 node exposes 192 GB of GPU memory in total.