Pyyan / Compare / CoreWeave vs RunPod vs Modal vs Vast.ai
CoreWeave vs RunPod vs Modal vs Vast.ai
4 of 5GPU Clouds · verified 13 Aug 2026
| Specification | CoreWeave | RunPod | Modal | Vast.ai |
|---|---|---|---|---|
| Summary | Large-scale training with capacity guarantees. | Best serverless and pay-per-request. | Code-first serverless, no idle cost. | Cheapest spot rates, no guarantees. |
| H100 ($/hr) | $— | $2.34 | — | — |
| B200 ($/hr) | $5.5 | $5.98 | — | — |
| Serverless | No | Yes | Yes | No |
| SLA | Yes | Yes | Yes | No |
| Category | GPU Clouds | GPU Clouds | GPU Clouds | GPU Clouds |
| Official | CoreWeave ↗ | RunPod ↗ | Modal ↗ | Vast.ai ↗ |
Highlighted rows are where these differ.
RunPod
- 81% cheaper than AWS for H100
- Serverless endpoints, scale to zero
Best for variable traffic.
Vast.ai
- No SLA, no security guarantees, no compliance
- Peer-to-peer marketplace
Best for fault-tolerant batch.