What is the cheapest cloud GPU right now?
Right now the cheapest verified GPU we track is a RTX A4000 at $0.15/hr on Hyperstack. The very lowest rates come from consumer cards and community tiers, which trade guaranteed uptime for a lower price. If your work can checkpoint and restart, that trade is usually worth it.
| GPU | VRAM | $/hr | Where | |
|---|---|---|---|---|
| RTX A4000 | 16 GB | $0.15 | Hyperstack on-demand | Rent → |
| Tesla V100 | 16 GB | $0.17 | DataCrunch on-demand | Rent → |
| RTX A5000 | 24 GB | $0.27 | RunPod secure cloud | Rent → |
| L4 | 24 GB | $0.39 | RunPod secure cloud | Rent → |
| A40 | 48 GB | $0.44 | RunPod secure cloud | Rent → |
| RTX 3090 | 24 GB | $0.46 | RunPod secure cloud | Rent → |
| RTX A6000 | 48 GB | $0.49 | RunPod secure cloud | Rent → |
| RTX 4090 | 24 GB | $0.69 | RunPod secure cloud | Rent → |
| RTX 6000 Ada | 48 GB | $0.77 | RunPod secure cloud | Rent → |
| L40 | 48 GB | $0.82 | RunPod secure cloud | Rent → |
Cheapest depends on what you need the GPU to do. For light inference, notebooks, or learning, an older consumer card like a 3090 or 4090 is the cheapest way to get real GPU time. For anything memory-hungry, the floor rises with the VRAM you require.
The headline number almost always comes from a spot or community listing. Those are fine for batch jobs and fault-tolerant training, but a live endpoint that cannot be interrupted should compare firm on-demand rates instead. We flag which tier each quote belongs to so the comparison stays honest.
The table lists the cheapest cards we verified today, sorted by price. Because it is rebuilt from the daily snapshot, the leader can change from one day to the next as providers adjust capacity.
Related questions
- What is the cheapest GPU for Stable Diffusion?
- What is the difference between spot and on-demand GPU pricing?
- Is 24GB of VRAM enough for running an LLM?
Numbers on this page come from today's verified snapshot. Full table on the homepage; method in the methodology.