CoreWeave
vs
Nebius
Both rent H100 to GB300 GPUs with spot pricing, managed Kubernetes and object storage. Nebius runs managed PostgreSQL and a model API, while CoreWeave rents bare-metal nodes with Slurm scheduling.
What's good about each
- Nvidia nodes from A100 up to GB300 NVL72 racks
- Bare-metal nodes under a managed Kubernetes control plane
- No charge for the Kubernetes control plane or Slurm
- Free egress and no object storage request fees
- Nvidia GPUs from L40S up to GB300 NVL72 racks
- Preemptible GPU instances at a lower hourly rate
- Managed Kubernetes and Slurm clusters at no extra charge
- Free outbound traffic from virtual machines
- Per-token inference API for open-weight models
Price comparison
Closest matching configuration on each side, monthly list price in USD.
| Service | CoreWeave | Nebius | Difference |
|---|---|---|---|
| Small VM | -- Not listed | $35.71 /mo 2 vCPU, 8 GB RAM (cpu-d3) | |
| Medium VM | -- Not listed | $71.42 /mo 4 vCPU, 16 GB RAM (cpu-d3) | |
| Large VM | -- Not listed | $142.85 /mo 8 vCPU, 32 GB RAM (cpu-d3) | |
| Block storage | $7.00 /mo 100 GB (Distributed File Storage) | $7.10 /mo 100 GB (Network SSD) | Within 5% |
| Object storage | $61.44 /mo 1 TB (Hot) | $19.42 /mo 1 TB (Standard) | 68% cheaper on Nebius |
| PostgreSQL | -- Not listed | $104.35 /mo 2 vCPU, 8 GB RAM, 50 GB storage (2vcpu-8gb) | |
| Egress | Free and unlimited | Free from VMs, billed from Object Storage |
Heads up: Figures take each provider's closest plan to a common setup, and vary by region and usage. Hover a service for the reference configuration. Verify before provisioning. More on how we price.
GPUs
CoreWeave offers 11 GPU models, while Nebius offers 8. Cheapest listed price per GPU per hour; GPUs both providers sell come first.
| GPU | CoreWeave | Nebius | Difference |
|---|---|---|---|
H100
2 configs · 1 config
|
$6.16 /GPU/hr On-demand | $3.85 /GPU/hr On-demand | 37% cheaper on Nebius |
|
B300
1 config · 1 config
|
$4.48 /GPU/hr Spot | $7.85 /GPU/hr On-demand | Different terms |
|
B200
2 configs · 1 config
|
$8.60 /GPU/hr On-demand | $7.15 /GPU/hr On-demand | 17% cheaper on Nebius |
|
RTX PRO 6000
3 configs · 1 config
|
$2.50 /GPU/hr On-demand | $1.80 /GPU/hr On-demand | 28% cheaper on Nebius |
|
H200
2 configs · 1 config
|
$6.30 /GPU/hr On-demand | $4.50 /GPU/hr On-demand | 29% cheaper on Nebius |
|
GB200
1 config · 1 config
|
$10.50 /GPU/hr On-demand | -- On request | |
|
GB300
1 config · 1 config
|
-- On request | -- On request | |
|
L40S
2 configs · 4 configs
|
$2.25 /GPU/hr On-demand | $1.55 /GPU/hr On-demand | 31% cheaper on Nebius |
|
A100
2 configs
|
$1.21 /GPU/hr Spot | -- | CoreWeave only |
|
GH200
1 config
|
$6.50 /GPU/hr On-demand | -- | CoreWeave only |
CoreWeave bills on-demand and spot; Nebius bills on-demand and spot. The line under each price names the billing type shown.
Services
Managed services each provider offers.
Compliance
Frameworks each provider states on its own site, each linked to its source.
Heads up: Each entry links to the provider's own statement. An attestation can cover one tier, one region or the platform alone rather than every service. Verify the scope before you rely on it. A dash means we found nothing to link to, not that the provider lacks it.
Company details
| Fact | CoreWeave | Nebius |
|---|---|---|
| Founded | 2017 | 2022 |
| Headquarters | USA | Netherlands |
| Locations | 55 | 10 |
| Example customers | OpenAI, Cohere, Mistral AI, NovelAI | Revolut, Sword Health, Brave, Voxel51, Recraft |
| Website | coreweave.com | nebius.com |
Alternatives to consider
Compare CoreWeave and Nebius against other cloud providers.
Our data for CoreWeave was last updated on Sept. 26, 2026, and for Nebius on Sept. 26, 2026.