Coreweave
Cloud computing provider specializing in high-performance computing.
Pricing Change History
11 changesRTX PRO 6000 Blackwell Spot price increased $9.24→$11.09/hr (NA). Also, GPU Spot pricing was revealed (was N/A) for HGX B300 ($35.84/hr), HGX B200 ($34.11/hr) and HGX H100 ($19.71/hr).
Quarterly roll-up · includes Jun 15 weekly, May 18 weekly
Reserved capacity section heading renamed to Reserved Compute Capacity, adding specificity to the section label.
Intel Emerald Rapids (8562Y+) General Purpose CPU spot price increased from $2.11 to $3.27/hour in both North America and Europe regions (55% increase).
Added NVIDIA HGX B300 GPU to on-demand instances in North America and Europe [8 GPUs, 270GB VRAM, 192 vCPUs, 4,096GB RAM, 61.44TB storage; Contact sales for pricing] NVIDIA B200 renamed to NVIDIA HGX B200 in both North America and Europe regions [Proper HGX product naming applied to Blackwell GPU; all specs and $68.80/hour pricing unchanged]
Bring Your Own IP networking option added with $4.00 one-time setup fee per IP Intel Ice Lake (6342) CPU instance added with 96 vCPUs, 384GB RAM, 19.2TB storage at $3.36/hour
AMD Turin (9655P) CPU instance added at $8.18/hour (General Purpose, 192 vCPUs, 1,536GB RAM, 7.68TB storage) Intel Emerald Rapids (8562Y+) CPU instance added at $5.31/hour (General Purpose, 64 vCPUs, 512GB RAM, 7.68TB storage)
Research Coreweave's pricing with the Pulse Agent
Ask anything about Coreweave's plans, packaging shifts, and how it compares across the market — answered live from verified data.
Ask the agentCurrent plans
Captured| Plan | Price |
|---|---|
| Managed Kubernetes (CKS) | Free |
| Direct Connect | $1,250 /mo |
| Distributed File Storage | $0.07 per instance hour |
| GPU Compute Spot | $19.71 per 8-GPU node per hour (NVIDIA HGX H100 reference) |
| AI Object Storage - Warm | $0.03 per instance hour + $0.03 per instance hour |
| Reserved Compute Capacity | Contact |
| |
| AI Object Storage - Archive | $0.01 per instance hour + $0.01 per instance hour |
| CPU Compute On-Demand | $7.78 per CPU node per hour (AMD Genoa 9454 reference) |
| CPU Compute Spot | $3.32 per CPU node per hour (AMD Genoa 9454 reference) |
| AI Object Storage - Cold | $0.02 per instance hour + $0.02 per instance hour |
| AI Object Storage - Hot | $0.06 per instance hour + $0.06 per instance hour |
| GPU Compute On-Demand | $49.24 per 8-GPU node per hour (NVIDIA HGX H100 reference) |
| GPU Inference (Single GPU) | $6.16 per GPU per hour (NVIDIA HGX H100 reference) |
Company Facts
How Coreweave compares with its peers
| Company | Entry price | Last change |
|---|---|---|
L Lambda Labs | $0.69 per GPU per hour | Jul 27, 2026 |
P Paper | $16 /user/mo, billed annually | Jul 6, 2026 |
R RunPod | $0.05 per GB/month | Oct 19, 2026 |
V Vast.ai | Usage-based | — |
H Huggingface | $9 /mo | Oct 16, 2026 |
T Together AI | $9 per GPU-hour (B200, promo) | Oct 26, 2026 |
Coreweave pricing FAQ
- How much does Coreweave cost?
- Coreweave pricing starts at $1,250/mo on Direct Connect. Coreweave lists 13 plans on its pricing page; the Managed Kubernetes (CKS) plan is free.
- Does Coreweave have a free plan?
- Yes. Coreweave's Managed Kubernetes (CKS) plan is free; paid plans start at $1,250/mo on Direct Connect.
- What pricing model does Coreweave use?
- Coreweave uses subscription + usage pricing across 13 plans, with a free tier.
- When did Coreweave last change its pricing?
- Jul 3, 2026: CoreWeave: GPU Spot pricing revealed; RTX PRO 6000 Spot price up.
- How often has Coreweave changed its pricing?
- Pulse has recorded 11 pricing changes for Coreweave since it started tracking on Nov 17, 2024; the pricing page was last captured on Jul 3, 2026.
- Who are Coreweave's pricing peers?
- Pulse compares Coreweave with Lambda Labs, Paper, RunPod, Vast.ai, Huggingface and Together AI.