GPU cloud · 6 GPU models tracked
Nebius
$1.55–$9.50per GPU-hour, the cheapest row to the dearest
On-demand and preemptible rates per GPU-hour. Preemptible capacity can be reclaimed mid-run. Commitment discounts are sold through sales and are not listed.
As of , Nebius lists 6 GPU models we track, from $1.55 per GPU-hour (NVIDIA L40S, on-demand, from) to $9.50 (NVIDIA B300, on-demand).
| Website | nebius.com |
|---|---|
| Pricing page | nebius.com/prices |
| 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: Nebius’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 | Nebius on-demand, per GPU-hour | Against the market (median rate card) |
|---|---|---|
| NVIDIA B300 | $9.50 captured Deploy ↗ | 6% above the median $8.99 · 26 providers 15 of 26 list lower · lowest Deep Infra $5.89 · compare |
| NVIDIA B200 | $8.50 captured Deploy ↗ | 22% above the median $6.94 · 40 providers 29 of 40 list lower · lowest Hostinger $4.50 from · compare |
| NVIDIA H200 | $5.40 captured Deploy ↗ | 18% above the median $4.56 · 56 providers 37 of 56 list lower · lowest Dataoorts $2.29 · compare |
| NVIDIA H100 SXM | $4.50 captured Deploy ↗ | 37% above the median $3.29 · 41 providers 31 of 41 list lower · lowest Cudo Compute $1.79 GPU only · compare |
| NVIDIA RTX PRO 6000 Blackwell | $1.80 captured Deploy ↗ | 18% below the median $2.20 · 42 providers 8 of 42 list lower · lowest Hostinger $0.65 from · compare |
| NVIDIA L40S | $1.55 from captured Deploy ↗ | 9% above the median $1.42 · 40 providers 20 of 40 list lower · lowest CloudRift $0.63 |
The same, in words
- Nebius lists the NVIDIA B300 from $9.50 per GPU-hour on-demand. The on-demand rate is unchanged since Oct 1, 2026 (7-day change: none). Against the market: 6% above the median rate card ($8.99 across 26 providers); 15 of the 26 providers we track list a lower on-demand rate, the lowest Deep Infra at $5.89 (Deep Infra’s rates, Nebius vs Deep Infra).
- Nebius lists the NVIDIA B200 from $8.50 per GPU-hour on-demand. The on-demand rate is unchanged since Oct 1, 2026 (7-day change: none). Against the market: 22% above the median rate card ($6.94 across 40 providers); 29 of the 40 providers we track list a lower on-demand rate, the lowest Hostinger at $4.50 (Hostinger’s rates, Nebius vs Hostinger).
- Nebius lists the NVIDIA H200 from $5.40 per GPU-hour on-demand. The on-demand rate is unchanged since Oct 1, 2026 (7-day change: none). Against the market: 18% above the median rate card ($4.56 across 56 providers); 37 of the 56 providers we track list a lower on-demand rate, the lowest Dataoorts at $2.29 (Dataoorts’s rates, Nebius vs Dataoorts).
- Nebius lists the NVIDIA H100 SXM from $4.50 per GPU-hour on-demand. The on-demand rate is unchanged since Oct 1, 2026 (7-day change: none). Against the market: 37% above the median rate card ($3.29 across 41 providers); 31 of the 41 providers we track list a lower on-demand rate, the lowest Cudo Compute at $1.79 (Cudo Compute’s rates, Nebius vs Cudo Compute).
- Nebius lists the NVIDIA RTX PRO 6000 Blackwell from $1.80 per GPU-hour on-demand. The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 18% below the median rate card ($2.20 across 42 providers); 8 of the 42 providers we track list a lower on-demand rate, the lowest Hostinger at $0.65 (Hostinger’s rates, Nebius vs Hostinger).
- Nebius lists the NVIDIA L40S from $1.55 per GPU-hour on-demand (from). The on-demand rate is unchanged since first read, Sep 7, 2026 (7-day change: none). Against the market: 9% above the median rate card ($1.42 across 40 providers); 20 of the 40 providers we track list a lower on-demand rate, the lowest CloudRift at $0.63 (CloudRift’s rates).
Dated pieces on Nebius: — Nebius H100: $3.85 on demand, from $0.79 spot, September 2026. Each is frozen as of its date.
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 L40Sfrom | $1.55 | captured | Deploy ↗ |
| NVIDIA RTX PRO 6000 Blackwell | $1.80 | captured | Deploy ↗ |
| NVIDIA H100 SXM | $4.50 | captured | Deploy ↗ |
| NVIDIA H200 | $5.40 | captured | Deploy ↗ |
| NVIDIA B200 | $8.50 | captured | Deploy ↗ |
| NVIDIA B300 | $9.50 | captured | Deploy ↗ |
Spot / interruptible
| GPU · what is sold | $ / GPU-hour | Read | Source |
|---|---|---|---|
| NVIDIA L40Sfrom | $0.74 | stoppedlast confirmed | |
| NVIDIA H200from | $0.79 | stoppedlast confirmed | |
| NVIDIA H100 SXMfrom | $0.79 | stoppedlast confirmed | |
| NVIDIA RTX PRO 6000 Blackwellfrom | $0.79 | stoppedlast confirmed | |
| NVIDIA B300from | $0.99 | stoppedlast confirmed | |
| NVIDIA B200from | $0.99 | stoppedlast confirmed |
Ask an AI agent about Nebius’s rates: Claude, Cursor, VS Code and ChatGPT can read them through the GPU Economy MCP server - free, no key, every answer citing the page and read time.
What else Nebius bills
The GPU-hour is not the whole bill. What Nebius'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 Nebius states | Source |
|---|---|---|
| In the rate | Preset vCPU and RAM appear inside the GPU-hour for H100/H200/B200/B300/RTX PRO; L40S bills CPU and RAM separately; disks extra our reading | source checked |
| Data out | free: egress and ingress traffic not charged | source checked |
| Storage | Network SSD disk $0.071/GiB-month (non-replicated $0.053; shared filesystem $0.08) | source checked |
| While stopped | Stopped VM: GPU, vCPU, RAM not charged; disks and filesystems keep billing | source checked |
| Billing step | Per second (priced per hour, charged proportionally); no minimum charge published | source checked |
Every rate on this page as data: /providers/nebius/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/nebius, an iframe, no script).
Compare Nebius with
Every provider that lists at least two of the same GPUs on-demand, most shared GPUs first; each opens the side-by-side page.
- Amazon Web Services 6 shared GPUs
- GPU.ai 6 shared GPUs
- Lium 6 shared GPUs
- Modal 6 shared GPUs
- Oracle Cloud (OCI) 6 shared GPUs
- RunPod 6 shared GPUs
- Verda 6 shared GPUs
- CoreWeave 5 shared GPUs
- EmpirioLabs 5 shared GPUs
- HexGrid 5 shared GPUs
- Hyperstack 5 shared GPUs
- Runware 5 shared GPUs
- Vast.ai 5 shared GPUs
- Barrack AI 4 shared GPUs
- Beam 4 shared GPUs
- Cerebrium 4 shared GPUs
- Civo 4 shared GPUs
- Crusoe 4 shared GPUs
- Daytona 4 shared GPUs
- E2E Networks 4 shared GPUs
- fal 4 shared GPUs
- FarmGPU 4 shared GPUs
- FPT AI Factory 4 shared GPUs
- Gcore 4 shared GPUs
- Geodd 4 shared GPUs
- Hugging Face 4 shared GPUs
- Koyeb 4 shared GPUs
- Lyceum 4 shared GPUs
- OpenRelay 4 shared GPUs
- Seeweb 4 shared GPUs
- Sesterce 4 shared GPUs
- Together AI 4 shared GPUs
- Yotta Labs 4 shared GPUs
- Bitdeer AI 3 shared GPUs
- Dataoorts 3 shared GPUs
- Deep Infra 3 shared GPUs
- Fireworks AI 3 shared GPUs
- Google Cloud 3 shared GPUs
- Hikube 3 shared GPUs
- Hostinger 3 shared GPUs
- HPC-AI.COM 3 shared GPUs
- Hyperbolic 3 shared GPUs
- Jarvislabs 3 shared GPUs
- Limestone Networks 3 shared GPUs
- Massed Compute 3 shared GPUs
- Microsoft Azure 3 shared GPUs
- Neysa 3 shared GPUs
- Northflank 3 shared GPUs
- packet.ai 3 shared GPUs
- Paperspace 3 shared GPUs
- Tata Communications 3 shared GPUs
- Airon 2 shared GPUs
- Atlas Cloud 2 shared GPUs
- Canopy Wave 2 shared GPUs
- Cudo Compute 2 shared GPUs
- Elice Cloud 2 shared GPUs
- Enverge 2 shared GPUs
- FlexAI 2 shared GPUs
- GMI Cloud 2 shared GPUs
- Lambda 2 shared GPUs
- Lightning AI 2 shared GPUs
- OVHcloud 2 shared GPUs
- QuantaCloud 2 shared GPUs
- Selectel 2 shared GPUs
- Servercore 2 shared GPUs
- STACKIT 2 shared GPUs
- Taiga Cloud (Northern Data Group) 2 shared GPUs
- VESSL AI 2 shared GPUs
Data updated: