What does a PetaFLOP cost right now?
Live GPU rental prices from 48 providers, normalized per GPU-hour and ranked by what you actually get: dense FLOPS, VRAM or memory bandwidth per dollar.
Find the sweet spot
Cheapest offer for each GPU: price vs. dense FP16 / BF16 throughput. Up and to the left is better; the line connects GPUs nothing else beats on both.
Best value by GPU
Ranked by $ / PFLOP·h at each GPU's cheapest matching offer. Click a column header to re-sort, or a GPU to compare every provider.
- 1
L40S$1.05/PFLOP·h$0.38/GPU·h at Lium · 0.362 PFBest - 2V100 SXM2 16GB$1.08/PFLOP·h$0.14/GPU·h at GPU.ai 🇺🇸 · 0.125 PFGreat +3%
- 3RTX A4000$1.15/PFLOP·h$0.09/GPU·h at Vast.ai 🇺🇸 · 0.0767 PFGreat +9%
- 4RTX 3080 Ti$1.21/PFLOP·h$0.08/GPU·h at Vast.ai 🇺🇸 · 0.0682 PFGreat +15%
- 5RTX 3090$1.27/PFLOP·h$0.09/GPU·h at Akash Network · 0.071 PFGreat +21%
- 6RTX 4070 SUPER$1.34/PFLOP·h$0.09/GPU·h at Vast.ai 🇺🇸 · 0.071 PFGreat +27%
- 7RTX 2080 Ti$1.42/PFLOP·h$0.08/GPU·h at Vast.ai 🇺🇸 · 0.0538 PFGood +35%
- 8RTX A5000$1.44/PFLOP·h$0.16/GPU·h at RunPod 🇺🇸 · 0.111 PFGood +37%
- 9T4$1.69/PFLOP·h$0.11/GPU·h at Akash Network · 0.065 PFGood +61%
- 10V100 PCIe$1.70/PFLOP·h$0.19/GPU·h at RunPod 🇺🇸 · 0.112 PFGood +62%
- 11RTX 4070 Ti$1.75/PFLOP·h$0.14/GPU·h at Vast.ai 🇺🇸 · 0.0802 PFGood +67%
- 12RTX 5080$1.78/PFLOP·h$0.20/GPU·h at Vast.ai 🇺🇸 · 0.113 PFGood +69%
- 13RTX 4080 SUPER$1.80/PFLOP·h$0.19/GPU·h at Vast.ai 🇺🇸 · 0.104 PFGood +72%
- 14L40$1.82/PFLOP·h$0.33/GPU·h at Lium · 0.181 PFGood +74%
- 15RTX 5070 Ti$1.83/PFLOP·h$0.16/GPU·h at Vast.ai 🇺🇸 · 0.0879 PFGood +74%
| # | |||||||
|---|---|---|---|---|---|---|---|
| 1 | 48 GB | 0.9 TB/s | 0.362 PF |  $0.38/GPU·hat Lium median $1.25 across 26 | $1.05/PFLOP·hBest | 26 → | |
| 2 | 16 GB | 0.9 TB/s | 0.125 PF |  $0.14/GPU·hat GPU.ai 🇺🇸 median $0.81 across 10 | $1.08/PFLOP·hGreat +3% | 10 → | |
| 3 | 16 GB | 0.4 TB/s | 0.0767 PF |  $0.09/GPU·hat Vast.ai 🇺🇸 median $0.17 across 7 | $1.15/PFLOP·hGreat +9% | 7 → | |
| 4 | 12 GB | 0.9 TB/s | 0.0682 PF | $0.08/GPU·hat Vast.ai 🇺🇸 median $0.18 across 3 | $1.21/PFLOP·hGreat +15% | 3 → | |
| 5 | 24 GB | 0.9 TB/s | 0.071 PF |  $0.09/GPU·hat Akash Network median $0.21 across 5 | $1.27/PFLOP·hGreat +21% | 5 → | |
| 6 | 12 GB | 0.5 TB/s | 0.071 PF | $0.09/GPU·hat Vast.ai 🇺🇸 | $1.34/PFLOP·hGreat +27% | 1 → | |
| 7 | 11 GB | 0.6 TB/s | 0.0538 PF | $0.08/GPU·hat Vast.ai 🇺🇸 | $1.42/PFLOP·hGood +35% | 1 → | |
| 8 | 24 GB | 0.8 TB/s | 0.111 PF |  $0.16/GPU·hat RunPod 🇺🇸 median $0.79 across 6 | $1.44/PFLOP·hGood +37% | 6 → | |
| 9 | 16 GB | 0.3 TB/s | 0.065 PF | $0.11/GPU·hat Akash Network median $0.48 across 8 | $1.69/PFLOP·hGood +61% | 8 → | |
| 10 | 16 GB | 0.9 TB/s | 0.112 PF | $0.19/GPU·hat RunPod 🇺🇸 median $2.46 across 4 | $1.70/PFLOP·hGood +62% | 4 → | |
| 11 | 12 GB | 0.5 TB/s | 0.0802 PF | $0.14/GPU·hat Vast.ai 🇺🇸 median $0.19 across 3 | $1.75/PFLOP·hGood +67% | 3 → | |
| 12 | 16 GB | 1 TB/s | 0.113 PF | $0.20/GPU·hat Vast.ai 🇺🇸 median $0.29 across 4 | $1.78/PFLOP·hGood +69% | 4 → | |
| 13 | 16 GB | 0.7 TB/s | 0.104 PF | $0.19/GPU·hat Vast.ai 🇺🇸 median $0.23 across 2 | $1.80/PFLOP·hGood +72% | 2 → | |
| 14 | 48 GB | 0.9 TB/s | 0.181 PF | $0.33/GPU·hat Lium median $0.86 across 8 | $1.82/PFLOP·hGood +74% | 8 → | |
| 15 | 16 GB | 0.9 TB/s | 0.0879 PF | $0.16/GPU·hat Vast.ai 🇺🇸 median $0.21 across 2 | $1.83/PFLOP·hGood +74% | 2 → |
Longer, greener bars mean more compute per dollar. The label shows how much more each GPU costs per unit than the best one. Click a GPU to compare every provider.
Data sources · 44/51 healthy
- Vast.ai347 offers
- RunPod66 offers
- Verda92 offers
- GPU.ai58 offers
- Shadeform187 offers
- Vultr4 offers
- AWS pricing114 offers
- Lambda28 offers
- Nebius7 offers
- Hyperstack30 offers
- DigitalOcean30 offers
- Together AIfailed
- Scalewayfailed
- OVHcloud37 offers
- Cudo Compute2 offers
- Hyperbolic5 offers
- Oracle Cloud15 offers
- Crusoefailed
- Hot Aisle4 offers
- TensorWave6 offers
- Cirrascalefailed
- Seeweb44 offers
- Azure Retail API74 offers
- Massed Compute74 offers
- Lium85 offers
- Akash Network7 offers
- CoreWeave34 offers
- Oblivusfailed
- Thunder Computefailed
- Koyeb22 offers
- Akamai (Linode)13 offers
- EmpirioLabs AI44 offers
- Sesterce62 offers
- Hetznerfailed
- STACKIT10 offers
- T Cloud Public49 offers
- IONOS4 offers
- Gcore18 offers
- Exoscale22 offers
- UpCloud76 offers
- Leaseweb69 offers
- Infomaniak42 offers
- Leafcloud12 offers
- Civo87 offers
- Hivenet4 offers
- Arkane Cloud2 offers
- Latitude9 offers
- SkyPilot: aws145 offers
- SkyPilot: gcp76 offers
- SkyPilot: nebius52 offers
- SkyPilot: seeweb28 offers
Browse by provider
See which GPUs each cloud rents out and how its prices compare.
- RunPod 🇺🇸42 GPUs · $0.12
- GPU.ai 🇺🇸27 GPUs · $0.09
- Vast.ai 🇺🇸23 GPUs · $0.08
Sesterce 🇫🇷22 GPUs · $0.17- Lium 21 GPUs · $0.16
EmpirioLabs AI 🇺🇸18 GPUs · $0.25
AWS 🇺🇸16 GPUs · $0.53
Massed Compute 🇺🇸16 GPUs · $0.35
Oracle Cloud 🇺🇸15 GPUs · $1.28
Google Cloud 🇺🇸12 GPUs · $0.54
OVHcloud 🇫🇷12 GPUs · $0.38
Verda (DataCrunch) 🇫🇮12 GPUs · $0.24
Azure 🇺🇸11 GPUs · $0.53
Hyperstack 🇬🇧11 GPUs · $0.15
Lambda 🇺🇸11 GPUs · $0.69
Seeweb 🇮🇹11 GPUs · $0.33
CoreWeave 🇺🇸10 GPUs · $1.25
DigitalOcean 🇺🇸10 GPUs · $0.76
Koyeb 🇫🇷10 GPUs · $0.50
Leaseweb 🇳🇱8 GPUs · $0.22
Paperspace 🇺🇸8 GPUs · $0.80
T Cloud Public 🇩🇪8 GPUs · $0.51- Akash Network 7 GPUs · $0.09
Civo 🇬🇧7 GPUs · $1.09
Exoscale 🇨🇭6 GPUs · $0.74
Nebius 🇳🇱6 GPUs · $1.55
UpCloud 🇫🇮6 GPUs · $0.69
Infomaniak 🇨🇭5 GPUs · $0.16
Crusoe 🇺🇸4 GPUs · $1.10
Denvr 🇨🇦4 GPUs · $1.10
Gcore 🇱🇺4 GPUs · $1.40
Latitude.sh 🇺🇸4 GPUs · $0.74
Leafcloud 🇳🇱4 GPUs · $0.68
Scaleway 🇫🇷4 GPUs · $0.95
STACKIT 🇩🇪4 GPUs · $1.95
IMWT 3 GPUs · $0.49
TensorWave 🇺🇸3 GPUs · $1.71
Akamai (Linode) 🇺🇸2 GPUs · $0.52
Arkane Cloud 🇫🇷2 GPUs · $2.49
Cudo Compute 🇬🇧2 GPUs · $0.88- Horizon 2 GPUs · $1.95
Hyperbolic 🇺🇸2 GPUs · $3.49- Phyntec 2 GPUs · $0.40
Vultr 🇺🇸2 GPUs · $1.67
Hivenet 🇫🇷1 GPU · $0.84
Hot Aisle 🇺🇸1 GPU · $2.99
IONOS 🇩🇪1 GPU · $3.36
Voltage Park 🇺🇸1 GPU · $1.99
Browse GPU prices
Every GPU we track right now, with its lowest on-demand price per hour.
NVIDIA65
- H100 SXM$1.82
- L40S$0.38
- RTX PRO 6000 Blackwell$1.25
- A100 SXM 80GB$1.07
- H200 SXM$2.92
- RTX A6000$0.33
- A100 PCIe 80GB$0.87
- B200 SXM$5.43
- L4$0.27
- B300 SXM$5.60
- RTX 5090$0.41
- RTX 4090$0.34
- V100 SXM2 16GB$0.14
- A100 PCIe 40GB$0.60
- H100 PCIe$1.66
- T4$0.11
- A100 SXM 40GB$0.91
- RTX 6000 Ada$0.72
- H100 NVL$2.59
- RTX A4000$0.09
- A30$0.33
- RTX 5080$0.20
- RTX 5060 Ti$0.10
- L40$0.33
- RTX 5070 Ti$0.16
- RTX A5000$0.16
- A40$0.53
- A10$0.85
- RTX 3090$0.09
- H200 NVL$3.36
- RTX 4080$0.23
- RTX 4080 SUPER$0.19
- RTX 4070 SUPER$0.09
- V100 PCIe 32GB$2.34
- V100 PCIe$0.19
- Quadro RTX 6000$0.33
- RTX PRO 4500 Blackwell$0.72
- GB300$13.7
- RTX 4000 Ada$0.20
- RTX 4070$0.12
- RTX 3080 Ti$0.08
- A10G$1.01
- RTX 4070 Ti$0.14
- GB200$10.5
- RTX PRO 5000 Blackwell$0.80
- RTX PRO 4000 Blackwell$0.27
- RTX 3070$0.13
- V100S PCIe$0.90
- RTX PRO 6000 Blackwell Max-Q$1.64
- RTX 4060 Ti$0.10
- RTX 3080$0.12
- RTX 2080 Ti$0.08
- GH200$2.59
- A16$0.52
- V100 SXM2 32GB$3.90
- RTX A4500$0.19
- A2$0.16
- P100$1.28
- RTX 5000 Ada$0.49
- RTX 4000 SFF Ada$0.50
- RTX 2000 Ada$0.24
- RTX A2000$0.12
- RTX 3090 Ti$0.27
- RTX 3060 Ti$0.15
- Titan RTX$0.26
AMD5
Google TPUAWS Trainium3
Intel1
Popular comparisons
- H100 SXM vs H200 SXM
- H100 SXM vs A100 SXM 80GB
- H100 SXM vs B200 SXM
- H200 SXM vs B200 SXM
- B200 SXM vs B300 SXM
- GB200 vs B200 SXM
- H100 SXM vs Instinct MI300X
- H200 SXM vs Instinct MI300X
- Instinct MI300X vs Instinct MI325X
- B200 SXM vs Instinct MI355X
- H100 PCIe vs H100 SXM
- H100 NVL vs H100 SXM
- GH200 vs H100 SXM
- A100 SXM 80GB vs A100 PCIe 80GB
- A100 SXM 40GB vs A100 SXM 80GB
- L40S vs A100 SXM 80GB
- All comparisons →
GPU pricing questions
What is $ per PFLOP-hour?
It is the rental price of one GPU for one hour divided by its peak compute in PetaFLOPS at the precision you choose. It shows how much compute each dollar buys, so a cheap slow GPU and an expensive fast one can be compared directly. Lower is better.
Which GPU gives the most compute per dollar right now?
At dense FP16/BF16, the L40S currently leads at $1.05 per PFLOP-hour ($0.38 per GPU-hour at Lium). Rankings change with prices; the table above updates every 2 hours.
How much does an H100 cost per hour?
On-demand H100 SXM prices currently start at $1.82 per GPU-hour, with a median of $3.20 across 29 providers. Spot capacity starts at $0.74.
Why dense FLOPS instead of the numbers on NVIDIA's spec sheets?
Headline tensor numbers often assume 2:4 structured sparsity, which doubles them but is rarely used in practice. PetaPrice always uses dense throughput so GPUs and vendors are compared fairly.
Where does the price data come from?
Directly from the providers wherever they publish prices (public APIs, official price lists and pricing pages), with Shadeform's multi-cloud API and SkyPilot's catalogs filling gaps: 1360 offers from 48 providers, refreshed about every 2 hours. Prices are indicative list prices; confirm with the provider before booking.
What's the difference between on-demand, spot and reserved?
On-demand is billed by the hour with no commitment. Spot (interruptible) capacity is cheaper but can be reclaimed at any time. Reserved prices require a commitment, typically 1 to 3 years. PetaPrice never mixes them; the default view is on-demand.