Ranking 01
Cheapest
Lowest true cost per GPU-hour, in stock or stock unknown, newest reading per provider
NVIDIA · Hopper · Datacenter · 24 providers list itNVIDIA H100 NVL cloud prices
Rented today by 15 providers with a current on-demand offer, from $1.88 to $8.98 per GPU-hour, median $3.94. Every number below links to the page it was read from, with its read date.
Memory94GB HBM3
Bandwidth3.94TB/s
FP16 tensor—TFLOPS
FP32—TFLOPS
Power—W
Specifications: manufacturer figures.
Cheapest on-demand · read 00:32 UTC
$1.88/ GPU·h
Airon AI1× H100 NVL · 94 GB · Sweden- Median
- $3.94 15 providers, once each
- 30 days
- +18.2%
- 90 days
- +16.9%
Prices by rental mode · On-demand · Spot · ReservedH100 NVL prices today
Cheapest, median and spread among current offers. Medians count one floor per marketplace and need at least three offers.
Cheapest$1.88Airon AI
Median$3.94n = 36
Highest$8.98among current offers
Current offers5215 providers · last 45 days
Ever listed8522 providers on file
Where today's on-demand offers sitone dot per eligible offer (36) · the pool behind the tiles · read Nov 2, 2026 00:32 UTC
Provider offer34Marketplace floor2Median$3.94Middle half$3.75 – $6.58
36 eligible offers from 15 providers, binned every $0.20: from $1.88 (Airon AI) to $8.98 (STACKIT). 6 further marketplace listings sit above their marketplace's floor and are not drawn (one floor per marketplace, as in the tiles). Hover a dot for the provider, price, configuration, region and read date; click it to open the provider.
On-demand price history · 34 daysdaily median across all providers · 2 provider lines from their 30-day series · read Nov 2, 2026
Median · all providers$3.94Providers
Today $3.9430 days +18.2%34 days +16.9%Low $2.53 · Oct 5High $4.47 · Oct 31
Daily median on-demand price per GPU-hour across all providers, 34 days to Nov 2, 2026. Provider lines: the provider's cheapest current on-demand offer, its listed price over the last 30 days (read Nov 2, 2026). Drag to zoom, double-click to reset.
Today's podium · on-demand · readings of the last 45 days · reference Nov 2, 2026 00:32 UTCWho rents the H100 NVL best today
Cheapest true cost, best throughput per dollar, highest Trust Score. One provider per step, ten places, every row dated.
Ranking 02
Best value
Most throughput per dollar for LLM inference · Trust ≥ 60 first, then provisional · estimated
Runpod1× · 00:03 UTCValue 72$2.59
Massed Compute1× · 00:32 UTCValue 64$3.11- 4
Microsoft Azure1× · UAE +14Value 23$6.98
- 5Airon AI1× · SwedenValue 81prov.$1.88
Ranking 03
Most trusted
Highest Trust Score among providers with an eligible offer · unscored providers follow, cheapest first
Vast.ai1× · CanadaTrust 78$2.34- 4Massed Compute1× · 00:32 UTCTrust 74$3.11
- 5
Spheron8× · CANADA-1Trust 53$2.75
36 eligible on-demand offers from 15 providers · median $3.94 (n = 36) · not ranked: 9 unavailable · readings of the last 45 days · reference Nov 2, 2026 · 00:32 UTCRankings are never sponsored. How we rank →
All offers · 143 readings on file · 24 providersEvery H100 NVL offer, with its receipt
Current offers ranked by true cost, each with its read date. Older readings stay listed below, in grey, and never enter a ranking.
| Provider | Configuration | Region | Stock | List | True cost | Read | |
|---|---|---|---|---|---|---|---|
 Clore.aiTrust 46 | 2× H100 NVL NVL80 vCPU · 630 GB RAM · 18.821 GB disk | US | Unavailable | $0.94 | $0.94 | 00:03 UTCfresh | Receipt |
| Clore.aiTrust 46marketplace flooroutlier · suspended | 2× H100 NVL NVL72 vCPU · 512 GB RAM · 488.081 GB disk | US | In stock | $1.04 | $1.04 | 00:03 UTCfresh | Receipt |
| Clore.aiTrust 46 | 2× H100 NVL NVL80 vCPU · 630 GB RAM · 16.716 GB disk | US | In stock | $1.04 | $1.04 | 00:03 UTCfresh | Receipt |
| Airon AI | 1× H100 NVL NVL | Sweden | Stock unknown | $1.88 | $1.88 | 00:24 UTCfresh | Receipt |
 GPU.ai | 8× H100 NVL NVL252 vCPU · 1,440 GB RAM · 100 GB disk | Canada | In stock | $2.06 | $2.06 | 00:20 UTCfresh | Receipt |
| Vast.aiTrust 78marketplace floor | 1× H100 NVL NVL40 vCPU · 314.7 GB RAM · 952 GB disk | Canada | In stock | $2.34 | $2.34 | 00:02 UTCfresh | Receipt |
| RunpodTrust 95 | 1× H100 NVL NVL | Runpod Community Cloud hosts | In stock | $2.59 | $2.59 | 00:03 UTCfresh | Receipt |
| GPU.ai | 1× H100 NVL NVL19 vCPU · 150 GB RAM · 100 GB disk | US | Unavailable | $2.59 | $2.59 | 00:20 UTCfresh | Receipt |
| GPU.ai | 1× H100 NVL NVL19 vCPU · 150 GB RAM · 100 GB disk | US | In stock | $2.59 | $2.59 | 00:20 UTCfresh | Receipt |
| GPU.ai | 1× H100 NVL NVL19 vCPU · 150 GB RAM · 100 GB disk | US | Unavailable | $2.59 | $2.59 | 00:20 UTCfresh | Receipt |
 GPURento | 1× H100 NVL NVL | Germany · France | Stock unknown | $2.59 | $2.59 | 00:25 UTCfresh | Receipt |
| Vast.aiTrust 78 | 2× H100 NVL NVL56 vCPU · 377.9 GB RAM · 5,610.2 GB disk | Bulgaria | In stock | $2.61 | $2.61 | 00:02 UTCfresh | Receipt |
| Vast.aiTrust 78 | 1× H100 NVL NVL28 vCPU · 188.9 GB RAM · 2,805.1 GB disk | Bulgaria | In stock | $2.62 | $2.62 | 00:02 UTCfresh | Receipt |
| 8× H100 NVL NVL252 vCPU · 1,440 GB RAM · 6,600 GB disk | CANADA-1 | In stock | $2.75 | $2.75 | 00:19 UTCfresh | Receipt | |
| 8× H100 NVL NVL252 vCPU · 1,440 GB RAM · 6,600 GB disk | CANADA-1 | In stock | $2.75 | $2.75 | 00:19 UTCfresh | Receipt | |
| Vast.aiTrust 78 | 1× H100 NVL NVL64 vCPU · 125.8 GB RAM · 187.5 GB disk | Japan | In stock | $2.76 | $2.76 | 00:02 UTCfresh | Receipt |
| Massed ComputeTrust 74 | 1× H100 NVL NVL | — | Stock unknown | $3.11 | $3.11 | 00:32 UTCfresh | Receipt |
| RunpodTrust 95 | 1× H100 NVL NVL | Runpod Secure Cloud datacenters | Unavailable | $3.19 | $3.19 | 00:03 UTCfresh | Receipt |
| GPU.ai | 1× H100 NVL NVL16 vCPU · 180 GB RAM · 100 GB disk | US | Unavailable | $3.19 | $3.19 | 00:20 UTCfresh | Receipt |
| GPU.ai | 1× H100 NVL NVL16 vCPU · 180 GB RAM · 100 GB disk | US | In stock | $3.19 | $3.19 | 00:20 UTCfresh | Receipt |
| GPU.ai | 1× H100 NVL NVL16 vCPU · 180 GB RAM · 100 GB disk | US | Unavailable | $3.19 | $3.19 | 00:20 UTCfresh | Receipt |
| Vast.aiTrust 78 | 1× H100 NVL NVL40 vCPU · 314.7 GB RAM · 858.2 GB disk | US | In stock | $3.30 | $3.30 | 00:03 UTCfresh | Receipt |
 GPUNETresold from other clouds (the upstream is named in each offer) | 8× H100 NVL NVL140 vCPU · 1,500 GB RAM · 10,000 GB disk | US | Unavailable | $3.32 | $3.32 | 00:24 UTCfresh | Receipt |
| GPUNETresold from other clouds (the upstream is named in each offer) | 2× H100 NVL NVL40 vCPU · 256 GB RAM · 2,500 GB disk | US | Unavailable | $3.32 | $3.32 | 00:24 UTCfresh | Receipt |
| GPUNETresold from other clouds (the upstream is named in each offer) | 4× H100 NVL NVL80 vCPU · 512 GB RAM · 5,000 GB disk | US | Unavailable | $3.32 | $3.32 | 00:24 UTCfresh | Receipt |
52 fresh offers · the 25 cheapest shown here, all of them in the comparison tool · 33 older readings of the same provider replaced by a newer one · prices per GPU-hour · true cost includes the add-ons the provider publishesOpen in the comparison tool →
Value · what the hour buysPrice per unit, and what fits
The same hour, divided by what the card gives you; then the smallest real configuration for the most downloaded open models.
| Unit | At the cheapest | At the median |
|---|---|---|
| per GB of VRAM · hour94 GB | $0.020 | $0.042 |
| per TB/s of bandwidth · hour3.94 TB/s | $0.477 | $1.001 |
| Model | Needs | Smallest config | $ / month |
|---|---|---|---|
| all MiniLM L6 v2Sentence Transformers | 1 GBFP8 | 1× H100 NVLat Airon AI | $1,372 |
| paraphrase multilingual MiniLM L12 v2Sentence Transformers | 1 GBFP8 | 1× H100 NVLat Airon AI | $1,372 |
| bge-m3BAAI | 1 GBFP8 | 1× H100 NVLat Airon AI | $1,372 |
| Qwen3 0.6BQwen | 1 GBFP8 | 1× H100 NVLat Airon AI | $1,372 |
| all-mpnet-base-v2Sentence Transformers | 1 GBFP8 | 1× H100 NVLat Airon AI | $1,372 |
| bge-reranker-v2-m3BAAI | 1 GBFP8 | 1× H100 NVLat Airon AI | $1,372 |
Monthly = 730 hours at the cheapest current on-demand price for that GPU count · models sized at FP8 when available, weights plus runtime overheadSize any model →
H100 NVL · comparisons and collectionsCompare the H100 NVL side by side
Questions about renting the H100 NVL
How much does the H100 NVL cost per hour?
Today the H100 NVL rents from $1.88 per GPU-hour on-demand, with a median of $3.94 across 15 providers, each counted once at its cheapest current offer (36 current offers in all, readings of the last 45 days). Prices depend on the provider, the number of GPUs in the machine, the region and whether you rent on-demand, spot or reserved; true cost adds the mandatory add-ons each provider publishes.
Which cloud providers offer the H100 NVL?
16 providers have a current H100 NVL offer in stock. Cheapest first: Clore Ai ($1.04), Airon Ai ($1.88), Gpu Ai ($2.06), Vast Ai ($2.34), Runpod ($2.59), Gpurento ($2.59). Older readings from other providers stay listed on this page, in grey, and are never ranked.
How much memory does the H100 NVL have?
94 GB of HBM3 at 3.94 TB/s. At FP8 that fits all MiniLM L6 v2 (1 GB), paraphrase multilingual MiniLM L12 v2 (1 GB), bge-m3 (1 GB), Qwen3 0.6B (1 GB), all-mpnet-base-v2 (1 GB) on a single card; larger models need two to eight GPUs or a different card.
Where is the H100 NVL cheapest right now?
The lowest true cost among eligible offers is $1.88 per GPU-hour at Airon AI for a 1-GPU machine in Sweden, read within the ranking window. Offers marked unavailable, older than 45 days, superseded by a newer reading, resold or outside the outlier band are listed but not ranked.
Is the H100 NVL cheaper on spot or reserved?
Spot (interruptible) capacity starts at $0.63 per GPU-hour; reserved terms start at $1.99. Spot suits jobs that checkpoint; reserved suits machines that run all month.