RunPod review: GPU cloud prices, pods and serverless
Our hands-on RunPod review: what it costs per GPU-hour, Secure vs Community Cloud, serverless endpoints, and who it suits.
Updated · Tested on: Not yet tested
We're still testing this product. Below are its current plans and prices; our rating and verdict come after hands-on testing.
Pros
- Wide GPU range on one page, from RTX 4090 to B200
- Community Cloud is much cheaper than Secure Cloud (RTX 4090: $0.34/hr vs $0.89/hr)
- Serverless GPU endpoints alongside regular pods
Cons
- Secure Cloud prices are above marketplace rivals for the same GPU
RunPod plans and prices
| Plan | Price | Specs / includes | Notes |
|---|---|---|---|
| RTX 4090 24GB (Secure Cloud) | $0.89/hr | 6 vCPU · 41 GB RAM | Secure Cloud, per GPU. Community Cloud price for the same GPU is $0.34/hr. Page default view is Secure Cloud; toggle switches to Community Cloud. |
| L40S 48GB (Secure Cloud) | $1.09/hr | 16 vCPU · 94 GB RAM | Secure Cloud, per GPU. Community Cloud $0.79/hr. |
| A100 80GB PCIe (Secure Cloud) | $1.79/hr | A100 PCIe · 8 vCPU · 117 GB RAM | Secure Cloud, per GPU. Community Cloud $1.19/hr. A100 SXM 80GB (16 vCPU, 125 GB RAM) is also $1.79/hr Secure, $1.39/hr Community. |
| H100 80GB SXM (Secure Cloud) | $3.99/hr | H100 SXM · 20 vCPU · 125 GB RAM | Secure Cloud, per GPU. Community Cloud $2.69/hr. H100 PCIe 80GB (16 vCPU, 188 GB RAM) is $2.89/hr Secure, $1.99/hr Community. H100 NVL 94GB $3.19/hr Secure, $2.59/hr Community. |
| H200 141GB (Secure Cloud) | $5.29/hr | 24 vCPU · 276 GB RAM | Secure Cloud, per GPU. Community Cloud $3.59/hr. |
| B200 180GB (Secure Cloud) | $7.99/hr | 28 vCPU · 283 GB RAM | Secure Cloud, per GPU. Community Cloud $5.98/hr. |
| RTX 4090 24GB (Community Cloud) | $0.34/hr | — | Community Cloud: GPUs in vetted third-party data centres, cheaper than Secure Cloud. |
| RTX 5090 32GB (Secure Cloud) | $1.19/hr | 9 vCPU · 35 GB RAM | Secure Cloud, per GPU. Community Cloud $0.69/hr. |
Secure Cloud prices shown, with Community Cloud prices in the notes (cheaper, third-party data centres). Serverless GPUs are priced separately. RunPod's pricing page said 'Updated October 8, 2026'.
Source: RunPod pricing page. Prices in USD, before tax.
What RunPod is
[Overview: pods, Secure vs Community Cloud, serverless, templates.]
Pricing in practice
[What a real session cost us, including storage and idle time.]
Performance
[Time to a running pod, model download speed, tokens per second on our standard test.]
Ease of use
[Dashboard, templates, API, documentation.]
Support
[How support answered our test questions.]
RunPod alternatives
| Provider | From | Starting plan | Free tier | Our rating | Link |
|---|---|---|---|---|---|
| Vast.aiChecked | $0.49/hr | RTX 4090 24GB | No free credits | Not rated | VisitOfficial site |
| LambdaChecked | $2.29/hr | GH200 96GB (1x instance)64 vCPU · 432 GB RAM | No | Not rated | VisitOfficial site |
| TensorDockChecked (partly unverified) | $0.35/hr | RTX 4090 24GB | No ($5 minimum deposit) | Not rated | VisitOfficial site |
| ModalChecked | $1.10/hr | A10 | $30/month of free credits on the Starter plan | Not rated | VisitOfficial site |
How we score: our testing method. We don't accept payment for ratings.