GPU cloud · 6 GPU models tracked
Vultr
$1.49–$3.20per GPU-hour, the cheapest row to the dearest
Read from Vultr's own open plans endpoint, api.vultr.com/v2/plans-metal, which anyone can open without a key and which each row links beside the provider name; the name leads to Vultr's pricing page, which answers our reader 403 but opens in a browser. Each plan there is priced twice, on demand and preemptible, and says of each price whether it can be deployed that way; only a price it says you can deploy is shown here. Prices are for the whole machine, divided by its number of GPUs. Plans that sell a fraction of a GPU are not listed.
As of , Vultr lists 6 GPU models we track, from $1.49 per GPU-hour (NVIDIA A100 80GB, spot) to $3.20 (NVIDIA B200, spot).
| Website | www.vultr.com/ |
|---|---|
| Pricing page | www.vultr.com/pricing/ |
| Regions and tiers | No region or tier named in the rates below: one global rate card |
| Models tracked | 6 |
Price by GPU
As of , per GPU-hour: Vultr’s on-demand rate for each card it lists, a cheaper market it also sells, and where the on-demand rate stands among the providers we track.
| GPU | Vultr on-demand, per GPU-hour | Against the market (median rate card) |
|---|---|---|
| NVIDIA B200 | $3.20 spot · no on-demand rate captured Deploy ↗ | no on-demand rate here, so no position against on-demand rate cards |
| NVIDIA A100 80GB | $2.40 4-GPU node captured spot $1.49 Deploy ↗ | 20% above the median $2.00 · 58 providers 35 of 58 list lower · lowest CloudRift $1.05 |
| NVIDIA H100 80GB (form factor not stated) | $2.30 spot · no on-demand rate captured Deploy ↗ | no on-demand rate here, so no position against on-demand rate cards |
| AMD Instinct MI325X | $2.00 spot · no on-demand rate captured Deploy ↗ | no on-demand rate here, so no position against on-demand rate cards |
| AMD Instinct MI300X | $1.85 spot · no on-demand rate captured Deploy ↗ | no on-demand rate here, so no position against on-demand rate cards |
| NVIDIA L40S | $1.49 spot · no on-demand rate captured Deploy ↗ | no on-demand rate here, so no position against on-demand rate cards |
The same, in words
- Vultr lists the NVIDIA B200 from $3.20 spot (8-GPU node).
- Vultr lists the NVIDIA A100 80GB from $2.40 per GPU-hour on-demand (4-GPU node), $1.49 spot (8-GPU node). The on-demand rate is unchanged since first read, Sep 29, 2026 (7-day change: none). Against the market: 20% above the median rate card ($2.00 across 58 providers); 35 of the 58 providers we track list a lower on-demand rate, the lowest CloudRift at $1.05 (CloudRift’s rates).
- Vultr lists the NVIDIA H100 80GB (form factor not stated) from $2.30 spot (8-GPU node).
- Vultr lists the AMD Instinct MI325X from $2.00 spot (8-GPU node).
- Vultr lists the AMD Instinct MI300X from $1.85 spot (8-GPU node).
- Vultr lists the NVIDIA L40S from $1.49 spot (8-GPU node).
All listed prices
One table per price type, never mixed. Every rate normalized to one GPU for one hour; a price opens the page it was read from, and 7d marks a rate that moved in the past seven days. The day each figure was first listed is in this page’s data, below.
On-demand
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA A100 80GBAPI$9.59/node ÷ 4 | $2.40 | captured | Deploy ↗ |
Spot / interruptible
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA A100 80GBAPI$11.92/node ÷ 8 · −$0.91 (−38%) vs its on-demand $2.40 | $1.49 | captured | Deploy ↗ |
| NVIDIA L40SAPI$11.92/node ÷ 8 | $1.49 | captured | Deploy ↗ |
| AMD Instinct MI300XAPI$14.80/node ÷ 8 | $1.85 | captured | Deploy ↗ |
| AMD Instinct MI325XAPI$16.00/node ÷ 8 | $2.00 | captured | Deploy ↗ |
| NVIDIA H100 80GB (form factor not stated)API$18.40/node ÷ 8 | $2.30 | captured | Deploy ↗ |
| NVIDIA B200API$25.60/node ÷ 8 | $3.20 | captured | Deploy ↗ |
What else Vultr bills
The GPU-hour is not the whole bill. What Vultr's own pages say about the rest, read by hand and checked on ; each line links the page it was read from. "Not published" means we looked and the pages do not say. Every provider side by side.
| What | What Vultr states | Source |
|---|---|---|
| In the rate | The bare-metal GPU plan bundles CPU, RAM, local NVMe and a bandwidth allowance in one hourly_cost. www.vultr.com/pricing returned 403. our reading | source checked |
| Data out | Plan allowance (15,360 GB on the GPU bare-metal plans), then $0.01/GB. Inbound is not metered. | source checked |
| While stopped | A stopped instance keeps billing in full (CPU, RAM, storage, IP) until it is destroyed. | source checked |
| Billing step | Hourly, minimum one hour. Monthly cap of 730 h for Cloud GPU and 672 h for bare metal (our rows are bare metal). | source checked |
Every rate on this page as data: /providers/vultr/prices.json — the same fields as prices.json, narrowed to this provider. CC BY 4.0, no key, no rate limit. On your own page: embed this table (/embed/providers/vultr, an iframe, no script).
Data updated: