45 offerings from 18 providers across serverless products and GPU instances.
60 complete comparable prices; incomplete and additive prices remain visible but are not ranked as totals.
12 of 16 serverless offerings include a scale-to-zero operating mode.
Published configurations: 80 GB
| GPU | Provider | Plan | Price | Location |
|---|---|---|---|---|
| H100 80 GB VRAM/GPU · fixed | Baseten.co H100 | H100 standard scale to zero · dedicated 1 minute increments · bills deploying, scaling up, scaling down, making predictions |
Component price — not a total | unknown |
| Verified 1 Oct 2026
Official price source
(opens in new tab) | ||||
| 1× H100 80 GB VRAM/GPU · fixed | Beam.cloud H100 | Standard serverless scale to zero 1 ms increments · bills code running |
Component price — not a total | unknown |
| The Serverless section includes a “With committed spend” / “Book a call” option without a published amount; it was omitted. | ||||
| 1× H100 VRAM unknown · fixed | Cerebrium H100 | Standard scale-to-zero scale to zero 1 second increments |
Component price — not a total | unknown |
| Verified 1 Oct 2026
Official price source
(opens in new tab) Enterprise pricing is listed as Custom and was omitted. 1 geography evidence item
| ||||
| 1× H100 VRAM unknown · fixed | Cerebrium H100 | Standard scale to zero 1 second increments · bills active compute time |
Component price — not a total | EU compute EU-constrained Sweden |
| Verified 1 Oct 2026
Official price source
(opens in new tab) Enterprise pricing is listed as Custom and was omitted. 3 geography evidence items
| ||||
| 1× H100 80 GB VRAM/GPU · fixed | Fal.ai H100 | Always on always on 1 second increments · bills setup, idle, running, draining, terminating |
≈ EUR 3.963/hour | unknown |
| Scale to zero scale to zero 1 second increments · bills setup, idle, running, draining, terminating |
≈ EUR 3.963/hour | |||
| Verified 1 Oct 2026
Official price source
(opens in new tab) The source lists discounted “As low as” prices, but those custom deployment prices are omitted because the terms require sales contact. | ||||
| 2× H100 80 GB VRAM/GPU · fixed | Koyeb 2x H100 | always-on always on 1 second increments · bills running |
≈ EUR 4.403/hour | unknown |
| scale-to-zero scale to zero 1 second increments · bills running |
≈ EUR 4.403/hour | |||
| Verified 1 Oct 2026
Official price source
(opens in new tab) 1 geography evidence item
| ||||
| 4× H100 80 GB VRAM/GPU · fixed | Koyeb 4x H100 | always-on always on 1 second increments · bills running |
≈ EUR 8.807/hour | unknown |
| scale-to-zero scale to zero 1 second increments · bills running |
≈ EUR 8.807/hour | |||
| Verified 1 Oct 2026
Official price source
(opens in new tab) 1 geography evidence item
| ||||
| 8× H100 80 GB VRAM/GPU · fixed | Koyeb 8x H100 | always-on always on 1 second increments · bills running |
≈ EUR 17.61/hour | unknown |
| scale-to-zero scale to zero 1 second increments · bills running |
≈ EUR 17.61/hour | |||
| Verified 1 Oct 2026
Official price source
(opens in new tab) 1 geography evidence item
| ||||
| 1× H100 80 GB VRAM/GPU · fixed | Koyeb H100 | always-on always on 1 second increments · bills running |
≈ EUR 2.202/hour | unknown |
| scale-to-zero scale to zero 1 second increments · bills running |
≈ EUR 2.202/hour | |||
| Verified 1 Oct 2026
Official price source
(opens in new tab) 1 geography evidence item
| ||||
| 1× H100 VRAM unknown · fixed | Modal Nvidia H100 SXM5 | Broad region scale to zero none minimum · bills running |
Component price — not a total | unknown |
| Narrow region scale to zero none minimum · bills running |
Component price — not a total | |||
| Standard scale to zero none minimum · bills running |
Component price — not a total | |||
5 geography evidence items
| ||||
| 1× H100 80 GB VRAM/GPU · fixed | OVHcloud h100-1-gpu | always-on always on · dedicated 1 minute increments · bills SCALING, RUNNING |
≈ EUR 2.985/hour | unknown |
| scale-to-zero scale to zero · dedicated 1 minute increments · bills SCALING, RUNNING |
≈ EUR 2.985/hour | |||
| Verified 1 Oct 2026
Official price source
(opens in new tab) | ||||
| 1× H100 80 GB VRAM/GPU · fixed | Replicate Nvidia H100 GPU | Public models request metered bills request execution time |
≈ EUR 4.835/hour | unknown |
| | ||||
| 2× H100 80 GB VRAM/GPU · fixed | Replicate 2x Nvidia H100 GPU | Public models request metered · commitment bills request execution time |
≈ EUR 9.67/hour | unknown |
| | ||||
| 4× H100 80 GB VRAM/GPU · fixed | Replicate 4x Nvidia H100 GPU | Public models request metered · commitment bills request execution time |
≈ EUR 19.34/hour | unknown |
| | ||||
| 8× H100 80 GB VRAM/GPU · fixed | Replicate 8x Nvidia H100 GPU | Public models request metered · commitment bills request execution time |
≈ EUR 38.68/hour | unknown |
| | ||||
| 1× H100 80 GB VRAM/GPU · fixed | Runpod H100 | Flex scale to zero 1 second increments · bills initialization, execution, idle timeout |
≈ EUR 4.218/hour | unknown |
| Verified 1 Oct 2026
Official price source
(opens in new tab) Active workers have discounts available through sales inquiry; no public price is published, so Active variants were omitted. | ||||
| Provider | Instance | VRAM/GPU | vCPUs | RAM | Storage | Region | Variant | Price | Verified |
|---|---|---|---|---|---|---|---|---|---|
| OVHcloud | h100-380 H100 80 GB | 80 GB | 30 | 380 GB | 4040 GB | Gravelines FR · gra · EU | Gravelines on-demand standard |
| |
| Scaleway | H100-1-80G NVIDIA H100-PCIe | 80 GB | 24 | 240 GB | 3000 GB | Warsaw PL · pl-waw-2 · EU | PL-WAW-2 on-demand standard |
| |
| Scaleway | H100-1-80G NVIDIA H100-PCIe | 80 GB | 24 | 240 GB | 3000 GB | Paris FR · fr-par-2 · EU | FR-PAR-2 on-demand standard |
| |
| OVHcloud | h100-760 H100 80 GB | 80 GB | 60 | 760 GB | 4040 GB | Gravelines FR · gra · EU | Gravelines on-demand standard |
| |
| Scaleway | H100-2-80G NVIDIA H100-PCIe | 80 GB | 48 | 480 GB | 6000 GB | Warsaw PL · pl-waw-2 · EU | PL-WAW-2 on-demand standard |
| |
| Scaleway | H100-2-80G NVIDIA H100-PCIe | 80 GB | 48 | 480 GB | 6000 GB | Paris FR · fr-par-2 · EU | FR-PAR-2 on-demand standard |
| |
| Scaleway | H100-SXM-2-80G NVIDIA H100-SXM | 80 GB | 32 | 240 GB | 3200 GB | Paris FR · fr-par-2 · EU | FR-PAR-2 on-demand standard |
| |
| OVHcloud | h100-1520 H100 80 GB | 80 GB | 120 | 1520 GB | 4040 GB | Gravelines FR · gra · EU | Gravelines on-demand standard |
| |
| Scaleway | H100-SXM-4-80G NVIDIA H100-SXM | 80 GB | 64 | 480 GB | 6400 GB | Paris FR · fr-par-2 · EU | FR-PAR-2 on-demand standard |
| |
| Verda | 1× H100 SXM5 · 80 GB/GPU · 30 vCPU / 120 GB RAM H100 SXM5 | 80 GB | 30 | 120 GB | unknown | unknown | Spot spot |
≈ EUR 1.594/hour | |
| Verda | 1× H100 SXM5 · 80 GB/GPU · 32 vCPU / 170 GB RAM H100 SXM5 | 80 GB | 32 | 170 GB | unknown | unknown | Spot spot |
≈ EUR 1.594/hour | |
| Nebius | NVIDIA H100 · 1 GPU / 16 vCPU / 200 GiB NVIDIA H100 | 80 GB | 16 | 200 GB | unknown | Finland FI · eu-north1 · EU | eu-north1 preemptible spot |
≈ EUR 1.893/hour | |
| Lambda | 1× NVIDIA H100 PCIe NVIDIA H100 PCIe | 80 GB | 26 | 225 GB | 1024 GB | unknown | On-demand standard |
≈ EUR 2.897/hour | |
| Verda | 1× H100 SXM5 · 80 GB/GPU · 30 vCPU / 120 GB RAM H100 SXM5 | 80 GB | 30 | 120 GB | unknown | unknown | On-demand standard |
≈ EUR 3.197/hour | |
| Verda | 1× H100 SXM5 · 80 GB/GPU · 32 vCPU / 170 GB RAM H100 SXM5 | 80 GB | 32 | 170 GB | unknown | unknown | On-demand standard |
≈ EUR 3.197/hour | |
| Nebius | NVIDIA H100 · 1 GPU / 16 vCPU / 200 GiB NVIDIA H100 | 80 GB | 16 | 200 GB | unknown | Finland FI · eu-north1 · EU | eu-north1 on-demand standard |
≈ EUR 3.391/hour | |
| Lambda | 1× NVIDIA H100 SXM NVIDIA H100 SXM | 80 GB | 26 | 225 GB | 2816 GB | unknown | On-demand standard |
≈ EUR 3.778/hour | |
| DigitalOcean | NVIDIA H100 | 80 GB | 20 | 240 GB | 5120 GB | unknown | On-demand standard |
≈ EUR 3.884/hour | |
| Paperspace | H100 NVIDIA HGX H100 | 80 GB | 20 | 250 GB | unknown | New York City, New York US · ny2 | NY2 on-demand standard |
≈ EUR 5.24/hour | |
| Lambda | 2× NVIDIA H100 SXM NVIDIA H100 SXM | 80 GB | 52 | 450 GB | 5632 GB | unknown | On-demand standard |
≈ EUR 7.38/hour | |
| Lambda | 4× NVIDIA H100 SXM NVIDIA H100 SXM | 80 GB | 104 | 900 GB | 11264 GB | unknown | On-demand standard |
≈ EUR 14.41/hour | |
| Nebius | NVIDIA H100 · 8 GPU / 128 vCPU / 1600 GiB NVIDIA H100 | 80 GB | 128 | 1600 GB | unknown | Finland FI · eu-north1 · EU | eu-north1 preemptible spot |
≈ EUR 15.15/hour | |
| CoreWeave | NVIDIA HGX H100 | 80 GB | 128 | 2048 GB | 62914.56 GB | europe | Europe spot spot |
≈ EUR 17.18/hour | |
| CoreWeave | NVIDIA HGX H100 | 80 GB | 128 | 2048 GB | 62914.56 GB | north-america | North America spot spot |
≈ EUR 17.36/hour | |
| Nebius | NVIDIA H100 · 8 GPU / 128 vCPU / 1600 GiB NVIDIA H100 | 80 GB | 128 | 1600 GB | unknown | Finland FI · eu-north1 · EU | eu-north1 on-demand standard |
≈ EUR 27.12/hour | |
| Lambda | 8× NVIDIA H100 SXM NVIDIA H100 SXM | 80 GB | 208 | 1800 GB | 22528 GB | unknown | On-demand standard |
≈ EUR 28.11/hour | |
| DigitalOcean | NVIDIA H100 · 8 GPU NVIDIA H100 | 80 GB | 160 | 1920 GB | 40960 GB | unknown | On-demand standard |
≈ EUR 31.07/hour | |
| CoreWeave | NVIDIA HGX H100 | 80 GB | 128 | 2048 GB | 62914.56 GB | north-america | North America on-demand standard |
≈ EUR 43.36/hour | |
| CoreWeave | NVIDIA HGX H100 | 80 GB | 128 | 2048 GB | 62914.56 GB | europe | Europe on-demand standard |
≈ EUR 43.36/hour |
1 offering · 1 billing option
Dedicated Deployments
1 offering · 1 billing option
Serverless
2 offerings · 2 billing options
Serverless
1 offering · 4 billing options
On-demand GPU Instances
2 offerings · 2 billing options
GPU Droplets
1 offering · 2 billing options
Serverless
1 offering · 2 billing options
On-Demand Deployments
4 offerings · 8 billing options
Inference Endpoints
4 offerings · 8 billing options
Serverless Compute
5 offerings · 5 billing options
GPU Cloud Instances
1 offering · 3 billing options
Serverless
2 offerings · 4 billing options
GPU Virtual Machines
4 offerings · 5 billing options
AI Deploy · Public Cloud GPU
1 offering · 1 billing option
GPU Machines
8 offerings · 8 billing options
Private Models · Public Models
1 offering · 1 billing option
Serverless
4 offerings · 6 billing options
GPU Instances
2 offerings · 4 billing options
GPU Instances