AWS vs Vast GPU Pricing
Side-by-side GPU cloud pricing comparison with live data and real-time availability.
AWS:20 GPUs
Vast:51 GPUs
Shared GPUs:10
Updated:
Vast is cheaper for 10 of 10 shared GPU models.
AWS is the enterprise standard with unmatched ecosystem depth and compliance, while Vast.ai offers dramatically lower prices through its GPU marketplace — often 5-10x cheaper for comparable hardware.
GPU Pricing Comparison
L4Vast cheaper
Difference: $0.77 (96%)
L40SVast cheaper
Difference: $1.46 (78%)
RTX PRO 4500Vast cheaper
Difference: $2.23 (88%)
A100Vast cheaper
Difference: $2.54 (93%)
RTX PRO 6000Vast cheaper
Difference: $2.70 (80%)
A100-80GBVast cheaper
Difference: $2.99 (87%)
V100Vast cheaper
Difference: $3.88 (99%)
H100Vast cheaper
Difference: $5.55 (81%)
H200Vast cheaper
Difference: $7.11 (90%)
B200Vast cheaper
Difference: $9.61 (68%)
| GPU Model | AWS on-demand $/GPU/hr | Vast on-demand $/GPU/hr | Difference | Cheaper | AWS Visit | Vast Visit |
|---|---|---|---|---|---|---|
| L4 | $0.80/GPU/hr on-demand | $0.03/GPU/hr on-demand Available | $0.77 (96%) | Vast | Visit | Visit |
| L40S | $1.86/GPU/hr on-demand | $0.40/GPU/hr on-demand Available | $1.46 (78%) | Vast | Visit | Visit |
| RTX PRO 4500 | $2.52/GPU/hr on-demand | $0.29/GPU/hr on-demand Limited | $2.23 (88%) | Vast | Visit | Visit |
| A100 | $2.74/GPU/hr on-demand | $0.20/GPU/hr on-demand Available | $2.54 (93%) | Vast | Visit | Visit |
| RTX PRO 6000 | $3.36/GPU/hr on-demand | $0.66/GPU/hr on-demand Available | $2.70 (80%) | Vast | Visit | Visit |
| A100-80GB | $3.43/GPU/hr on-demand | $0.44/GPU/hr on-demand Available | $2.99 (87%) | Vast | Visit | Visit |
| V100 | $3.90/GPU/hr on-demand | $0.02/GPU/hr on-demand Available | $3.88 (99%) | Vast | Visit | Visit |
| H100 | $6.88/GPU/hr on-demand | $1.33/GPU/hr on-demand Available | $5.55 (81%) | Vast | Visit | Visit |
| H200 | $7.91/GPU/hr on-demand | $0.80/GPU/hr on-demand Available | $7.11 (90%) | Vast | Visit | Visit |
| B200 | $14.24/GPU/hr on-demand | $4.63/GPU/hr on-demand Limited | $9.61 (68%) | Vast | Visit | Visit |
| A10 | -- | $0.20/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| A10G | $1.01/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| A16 | -- | $0.11/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| B300 | $17.80/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| Inferentia | $0.23/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| Inferentia2 | $0.76/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| L40 | -- | $0.40/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Quadro RTX 6000 | -- | $0.13/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Quadro RTX 8000 | -- | $0.24/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Radeon Pro V520 | $0.38/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| RTX 2060 | -- | $0.01/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 2070 | -- | $0.05/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| RTX 2080 | -- | $0.74/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| RTX 2080 Ti | -- | $0.04/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3060 | -- | $0.02/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3060 Ti | -- | $0.03/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3070 | -- | $0.04/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3070 Ti | -- | $0.05/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3080 | -- | $0.03/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3080 Ti | -- | $0.05/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3090 | -- | $0.05/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 3090 Ti | -- | $0.12/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 4000 Ada | -- | $0.14/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| RTX 4060 | -- | $0.05/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 4060 Ti | -- | $0.04/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 4070 | -- | $0.07/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 4070 Ti | -- | $0.07/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 4080 | -- | $0.10/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 4090 | -- | $0.11/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 5060 | -- | $0.06/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| RTX 5060 Ti | -- | $0.04/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 5070 Ti | -- | $0.08/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 5080 | -- | $0.10/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 5090 | -- | $0.30/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX 6000 Ada | -- | $0.28/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX A2000 | -- | $0.03/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX A4000 | -- | $0.05/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX A4500 | -- | $0.08/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| RTX A5000 | -- | $0.07/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| RTX A6000 | -- | $0.29/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX PRO 4000 | -- | $0.14/GPU/hr on-demand Available | -- | -- | Visit | Visit |
| RTX PRO 5000 | -- | $0.54/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| T4 | $0.53/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| T4g | $0.42/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| Tesla P4 | -- | $0.02/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Tesla P40 | -- | $0.03/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Titan RTX | -- | $0.11/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Titan V | -- | $0.09/GPU/hr on-demand Limited | -- | -- | Visit | Visit |
| Trainium | $1.34/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
| Trainium2 | -- | -- | -- | -- | Visit | Visit |
| Virtex UltraScale+ (VU47P) | $1.98/GPU/hr on-demand | -- | -- | -- | Visit | Visit |
Egress Cost Comparison
AWS
$0.09/GB
Vast
Free
Exclusive GPUs
Only on AWS (9)
- A10G
- B300
- Inferentia
- Inferentia2
- Radeon Pro V520
- T4
- T4g
- Trainium
- Virtex UltraScale+ (VU47P)
Only on Vast (41)
- A10
- A16
- L40
- Quadro RTX 6000
- Quadro RTX 8000
- RTX 2060
- RTX 2070
- RTX 2080
- RTX 2080 Ti
- RTX 3060
- RTX 3060 Ti
- RTX 3070
- RTX 3070 Ti
- RTX 3080
- RTX 3080 Ti
- RTX 3090
- RTX 3090 Ti
- RTX 4000 Ada
- RTX 4060
- RTX 4060 Ti
- RTX 4070
- RTX 4070 Ti
- RTX 4080
- RTX 4090
- RTX 5060
- RTX 5060 Ti
- RTX 5070 Ti
- RTX 5080
- RTX 5090
- RTX 6000 Ada
- RTX A2000
- RTX A4000
- RTX A4500
- RTX A5000
- RTX A6000
- RTX PRO 4000
- RTX PRO 5000
- Tesla P4
- Tesla P40
- Titan RTX
- Titan V
Which Should You Choose?
Choose AWS if...
- You need enterprise compliance, SLAs, and support contracts
- Your workload depends on the AWS ecosystem (SageMaker, S3, EKS)
- You need guaranteed availability with reserved capacity
Choose Vast if...
- You want the absolute lowest GPU prices available anywhere
- You're running research or experiments where cost matters more than reliability
- You need access to consumer GPUs not available on AWS