Lowest observed rate
Nebius AI Cloud provides the lowest row in this exact evidence set at $3.9500 per GPU-hour. That is a list-price observation, not proof that capacity is rentable now.
Exact GPU provider comparison
CoreWeave vs Nebius AI Cloud B200 pricing
Compare CoreWeave vs Nebius AI Cloud B200 GPU prices: 5 official-source rates across 2 regions, $3.950–$8.600/GPU-hour. Estimate monthly cost.
This evidence snapshot is older than 18 hours. The calculator is paused until Guardian publishes a verified refresh; use the official evidence links below meanwhile.
Decision readout
What the current evidence says
Price dispersion
The highest row is 117.7% above the lowest. At 160 GPU-hours, selecting opposite ends of the observed range changes the compute line item by $744.00.
Capacity boundary
4 of 5 rates carry separate deployment-eligibility evidence. That signal is not live stock; unknown stays unknown and never inherits confidence from a fresh price.
on demand floor gap
Nebius AI Cloud's lowest observed on demand B200 rate is 16.9% below CoreWeave's floor ($7.1500 vs $8.6000/GPU-hour). At 160 GPU-hours, that floor-to-floor difference is $232.00. Same GPU model and purchase model; variant, region, node configuration, contract terms and evidenced availability can still differ.
Variant evidence: Nebius AI Cloud: NVLink 180GB · CoreWeave: not specified.
Deployability evidence coverage: Nebius AI Cloud 2/2 rows (100%) vs CoreWeave 0/1 (0%). This compares deployment eligibility, not live stock or actual rentable capacity.
This comparison holds the accelerator model constant at B200 and changes the provider. Nebius AI Cloud has the lowest observed row; CoreWeave contains the highest observed row. Before choosing, compare accelerator variant, region, GPU count per node, market type and separately evidenced capacity—not only the headline rate.
Side-specific evidence profile
What each provider contributes to this comparison
The profile below prevents a large catalog from silently overwhelming a smaller one. Counts, regions, access models and observed price bands are stated independently for every provider.
Official configuration set
Nebius AI Cloud
Nebius AI Cloud contributes 4 qualifying configurations and 4 distinct source records. Nebius AI Cloud's observed price band runs from $3.9500 to $7.1500 per GPU-hour, equivalent to $632.00–$1144.00 for 160 GPU-hours.
Nebius AI Cloud covers 2 named regions: me-west1, us-central1. The included node configurations range from 1 to 1 GPUs, with 2 on-demand and 2 spot records.
Nebius AI Cloud has independent deployment-eligibility evidence for 4 of 4 rows. This does not assert live stock. Its latest included price observation is 2026-08-13. GPU models in this exact comparison are B200.
- Configurations
- 4
- Distinct source keys
- 4
- Named regions
- 2
- Deployability evidenced
- 4/4
Official configuration set
CoreWeave
CoreWeave contributes 1 qualifying configuration and 1 distinct source record. CoreWeave's observed price band runs from $8.6000 to $8.6000 per GPU-hour, equivalent to $1376.00–$1376.00 for 160 GPU-hours.
CoreWeave covers 0 named regions: . The included node configurations range from 1 to 1 GPUs, with 1 on-demand and 0 spot records.
CoreWeave has independent deployment-eligibility evidence for 0 of 1 rows. This does not assert live stock. Its latest included price observation is 2026-08-13. GPU models in this exact comparison are B200.
- Configurations
- 1
- Distinct source keys
- 1
- Named regions
- 0
- Deployability evidenced
- 0/1
Transparent scenario calculator
Convert hourly rates into your budget
The calculation multiplies GPU count by hours per GPU and the selected normalized rate. It excludes storage, egress, support, taxes, reservations and idle orchestration overhead.
Configuration-level evidence
Rates included in this decision
Every row preserves the source configuration and observation date. Node rates are divided by GPU count to produce the normalized comparison. Equal normalized prices can still represent different regions, node sizes or access conditions.
| Provider / configuration | GPU | Region | Market | Node price | GPU-hour | 160h cost | Availability | Evidence |
|---|---|---|---|---|---|---|---|---|
| Nebius AI Cloudnebius:gpu-b200-sxm-a:me-west1:spot | B200NVLink 180GB · 1 GPU/node | me-west1 | spot | $3.9500 | $3.9500 | $632.00 | availableexact configuration · 2026-08-13 | Official source ↗Checked 2026-08-13 |
| Nebius AI Cloudnebius:gpu-b200-sxm:us-central1:spot | B200NVLink 180GB · 1 GPU/node | us-central1 | spot | $3.9500 | $3.9500 | $632.00 | availableexact configuration · 2026-08-13 | Official source ↗Checked 2026-08-13 |
| Nebius AI Cloudnebius:gpu-b200-sxm-a:me-west1:on_demand | B200NVLink 180GB · 1 GPU/node | me-west1 | on demand | $7.1500 | $7.1500 | $1144.00 | availableexact configuration · 2026-08-13 | Official source ↗Checked 2026-08-13 |
| Nebius AI Cloudnebius:gpu-b200-sxm:us-central1:on_demand | B200NVLink 180GB · 1 GPU/node | us-central1 | on demand | $7.1500 | $7.1500 | $1144.00 | availableexact configuration · 2026-08-13 | Official source ↗Checked 2026-08-13 |
| CoreWeavecoreweave:b200 | B200Variant not specified · 1 GPU/node | Global / unspecified | on demand | $68.8000 | $8.6000 | $1376.00 | Not evidencedunknown | Official source ↗Checked 2026-08-13 |
Comparison discipline
What this page can and cannot decide
It can compare list-price evidence
All included rows use fresh or verified official sources, USD pricing and a common per-GPU-hour unit. The page is regenerated when source evidence changes.
It cannot promise capacity
A fresh price is not inventory. Availability is shown only when a separate exact-configuration source supports it, with its own observation clock and limitations.
It exposes the omitted costs
Networking, storage, support, taxes, commitments, setup work and interruption risk can change total cost. Confirm those terms before purchasing.
It retires weak URLs
This page exists only while automated evidence and uniqueness thresholds pass. One failed cycle removes it from the index; three consecutive failed cycles retire the route.
Adjacent decisions
Continue with the closest evidence sets
These links are selected by shared provider or GPU model, so the next page changes one decision variable instead of sending you to a generic article.
provider model comparison
Lambda vs Nebius AI Cloud · B200
5 official-source rates · 2 regions · $3.950–$7.150/GPU-hour.
Open comparisonprovider model comparison
Nebius AI Cloud vs Verda · B200
5 official-source rates · 2 regions · $3.950–$7.150/GPU-hour.
Open comparisonprovider model
Nebius AI Cloud B200 pricing
4 official-source rates · 2 regions · $3.950–$7.150/GPU-hour.
Open comparisonprovider model comparison
Microsoft Azure vs Nebius AI Cloud · H100
185 official-source rates · 32 regions · $1.290–$23.326/GPU-hour.
Open comparisonprovider model comparison
Civo vs Nebius AI Cloud · H200
9 official-source rates · 4 regions · $2.450–$4.500/GPU-hour.
Open comparisonprovider model comparison
Crusoe Cloud vs Nebius AI Cloud · H200
9 official-source rates · 4 regions · $2.450–$4.500/GPU-hour.
Open comparison