GPU cloud prices, per GPU-hour
65 offers from 18 providers, 65 with a published price. Sorted by on-demand price per GPU-hour in US$; instance prices are divided by the number of GPUs. Prices checked 16 Sept 2026.
All GPUsH100 15A100 11L40S 10H200 7B200 7RTX PRO 6000 5B300 3L4 3MI300X 3GH200 1GB300 1L40 1Quadro RTX 6000 1RTX 4000 Ada 1
| # | Provider | Offer | On-demand | Spot | Committed | As published · source | Actions |
|---|---|---|---|---|---|---|---|
| 1 | Jarvislabs | L4 x1 1× L4 · 32 vCPU · 124 GB RAM | $0.44/GPU-hr | — | — | ||
| 2 | Akamai Cloud (Linode) | RTX 4000 Ada x1 1× RTX 4000 Ada Generation · 4 vCPU · 16 GB RAM · 500 GB disk | $0.52/GPU-hr | — | — | ||
| 3 | UpCloud | L4 x1 1× L4 · 8 vCPU · 64 GB RAM | $0.67/GPU-hr | — | — | ||
| 4 | OVHcloud | L4 x1 1× L4 · 22 vCPU · 90 GB RAM · 400 GB disk | $0.72/GPU-hr | — | — | ||
| 5 | Hyperstack | L40 x1 1× L40 | $1.00/GPU-hr | — | — | ||
| 6 | E2E Networks | L40S x1 1× L40S · 60 vCPU · 220 GB RAM | $1.20/GPU-hr | — | — | ||
| 7 | UpCloud | L40S x1 1× L40S · 8 vCPU · 64 GB RAM | $1.27/GPU-hr | — | — | ||
| 8 | Cyfuture | L40S x1 1× L40S · 16 vCPU · 256 GB RAM | $1.29/GPU-hr | — | — | ||
| 9 | OVHcloud | L40S x1 1× L40S · 15 vCPU · 90 GB RAM · 400 GB disk | $1.34/GPU-hr | — | — | ||
| 10 | Verda (formerly DataCrunch) | L40S x1 1× L40S | $1.41/GPU-hr | — | — | ||
| 11 | AceCloud | L40S x1 1× L40S · 16 vCPU · 64 GB RAM | $1.48/GPU-hr | — | — | ||
| 12 | Jarvislabs | A100 x1 1× A100 80GB · 16 vCPU · 112 GB RAM | $1.49/GPU-hr | — | — | ||
| 13 | Akamai Cloud (Linode) | Quadro RTX 6000 x1 1× Quadro RTX 6000 · 8 vCPU (dedicated) · 32 GB RAM · 640 GB disk | $1.50/GPU-hr | — | — | ||
| 14 | Crusoe | L40S x1 1× L40S | $1.50/GPU-hr | — | — | ||
| 15 | Nebius | L40S x1 1× L40S | $1.55/GPU-hr | — | — | ||
| 16 | Hyperstack | A100 x1 1× A100 80GB | $1.60/GPU-hr | — | — | ||
| 17 | Vultr | L40S x1 1× L40S | $1.67/GPU-hr | — | — | ||
| 18 | Verda (formerly DataCrunch) | A100 x1 1× A100 80GB | $1.73/GPU-hr | — | — | ||
| 19 | Nebius | RTX PRO 6000 x1 | $1.80/GPU-hr | — | — | ||
| 20 | Verda (formerly DataCrunch) | RTX PRO 6000 x1 | $1.84/GPU-hr | — | — | ||
| 21 | Hyperstack | RTX PRO 6000 x1 | $1.85/GPU-hr | — | — | ||
| 22 | Jarvislabs | RTX PRO 6000 x1 1× RTX PRO 6000 Blackwell Server Edition · 28 vCPU · 160 GB RAM | $1.89/GPU-hr | — | — | ||
| 23 | Krutrim Cloud | A100 x1 1× A100 80GB · 24 vCPU · 96 GB RAM | $1.97/GPU-hr | — | $1.02/GPU-hr 1-yr reserved | ||
| 24 | Voltage Park | H100 x1 1× H100 | $1.99/GPU-hr | — | — | ||
| 25 | Vultr | GH200 x1 | $1.99/GPU-hr | — | — | ||
| 26 | Crusoe | A100 80GB PCIe 1× A100 80GB | $2.00/GPU-hr | — | — | ||
| 27 | UpCloud | H100 x1 1× H100 · 12 vCPU · 240 GB RAM | $2.06/GPU-hr | $2.04/GPU-hr | — | ||
| 28 | Cyfuture | A100 x1 1× A100 80GB · 8 vCPU · 64 GB RAM | $2.06/GPU-hr | — | — | ||
| 29 | E2E Networks | A100 x1 1× A100 80GB · 16 vCPU · 115 GB RAM | $2.10/GPU-hr | — | — | ||
| 30 | Krutrim Cloud | H100 x1 1× H100 · 24 vCPU · 200 GB RAM | $2.22/GPU-hr | — | $1.80/GPU-hr 1-yr reserved | ||
| 31 | Crusoe | A100 x1 1× A100 80GB | $2.30/GPU-hr | — | — | ||
| 32 | Hyperstack | H100 x1 1× H100 | $2.50/GPU-hr | $2.00/GPU-hr | — | ||
| 33 | AceCloud | H100 x1 1× H100 · 26 vCPU · 250 GB RAM | $2.57/GPU-hr | — | — | ||
| 34 | OVHcloud | A100 x1 1× A100 80GB · 15 vCPU · 180 GB RAM · 300 GB disk | $2.63/GPU-hr | — | — | ||
| 35 | E2E Networks | H100 x1 1× H100 · 26 vCPU · 250 GB RAM | $2.66/GPU-hr | $0.73/GPU-hr | $1.62/GPU-hr Committed (3-mo) | ||
| 36 | Cyfuture | RTX PRO 6000 x1 1× RTX PRO 6000 Blackwell Server Edition · 16 vCPU · 128 GB RAM | $2.67/GPU-hr | — | — | ||
| 37 | Jarvislabs | H100 x1 1× H100 · 16 vCPU · 200 GB RAM | $2.69/GPU-hr | — | — | ||
| 38 | Lambda | A100 A100 80GB · 240 vCPU | $2.79/GPU-hr | — | — | ||
| 39 | Vultr | A100 | $2.80/GPU-hr | — | — | ||
| 40 | Cyfuture | MI300X x1 1× MI300X · 16 vCPU · 256 GB RAM | $2.86/GPU-hr | — | — | ||
| 41 | Scaleway | H100 x1 1× H100 · 24 vCPU · 240 GB RAM | $3.14/GPU-hr | — | — | ||
| 42 | AceCloud | H200 x1 1× H200 · 16 vCPU · 128 GB RAM | $3.18/GPU-hr | — | — | ||
| 43 | Verda (formerly DataCrunch) | H100 x1 1× H100 | $3.27/GPU-hr | $1.63/GPU-hr | — | ||
| 44 | Crusoe | MI300X x1 1× MI300X | $3.45/GPU-hr | — | — | ||
| 45 | Oracle Cloud Infrastructure | L40S x1 1× L40S | $3.50/GPU-hr | — | — | ||
| 46 | Nebius | H100 x1 1× H100 | $3.85/GPU-hr | $2.15/GPU-hr | — | ||
| 47 | E2E Networks | H200 x1 1× H200 · 30 vCPU · 375 GB RAM | $3.99/GPU-hr | — | — | ||
| 48 | Hyperstack | H200 x1 1× H200 | $3.99/GPU-hr | — | — | ||
| 49 | Jarvislabs | H200 x1 1× H200 · 28 vCPU · 300 GB RAM | $3.99/GPU-hr | — | — | ||
| 50 | Lambda | H100 H100 · 208 vCPU | $3.99/GPU-hr | — | — | ||
| 51 | Verda (formerly DataCrunch) | H200 x1 1× H200 | $4.24/GPU-hr | — | — | ||
| 52 | Nebius | H200 x1 1× H200 | $4.50/GPU-hr | $2.45/GPU-hr | — | ||
| 53 | RunPod | H100 serverless 1× H100 | $4.79/GPU-hr | — | — | ||
| 54 | UpCloud | B200 x1 1× B200 · 12 vCPU · 240 GB RAM | $5.17/GPU-hr | — | — | ||
| 55 | Hyperstack | B200 x1 1× B200 | $6.00/GPU-hr | — | — | ||
| 56 | Verda (formerly DataCrunch) | B200 x1 1× B200 | $6.29/GPU-hr | — | — | ||
| 57 | E2E Networks | B200 x1 1× B200 · 32 vCPU · 400 GB RAM | $6.99/GPU-hr | — | — | ||
| 58 | Nebius | B200 x1 1× B200 | $7.15/GPU-hr | $3.95/GPU-hr | — | ||
| 59 | Hyperstack | B300 x1 1× B300 | $7.40/GPU-hr | — | — | ||
| 60 | Verda (formerly DataCrunch) | B300 x1 1× B300 | $7.65/GPU-hr | — | — | ||
| 61 | Nebius | B300 x1 1× B300 | $7.85/GPU-hr | $4.30/GPU-hr | — | ||
| 62 | Verda (formerly DataCrunch) | GB300 x1 1× GB300 Grace Blackwell Ultra Superchip · 32 vCPU · 225 GB RAM | $8.80/GPU-hr | $4.40/GPU-hr | — | ||
| 63 | Oracle Cloud Infrastructure | H100 x1 1× H100 | $9.96/GPU-hr | — | — | ||
| 64 | Oracle Cloud Infrastructure | H200 x1 1× H200 | $10.00/GPU-hr | — | — | ||
| 65 | Oracle Cloud Infrastructure | B200 x1 1× B200 | $14.00/GPU-hr | — | — |
How to read this table
- On-demand is the pay-as-you-go list price. Intro offers are compared at their renewal price.
- Spot can be interrupted; Committed needs a term (shown under the price).
- As published is the provider’s own figure, currency and unit, before conversion.
- “Contact for pricing” means the provider does not publish a price. Those rows always sort last.
Before you buy
Availability changes hourly and regions differ. Confirm the price on the provider’s page (linked in every row), and check the interconnect (NVLink / InfiniBand) if you train across GPUs. See how we compare.
65 priced GPU offers compared