CoreWeave Pricing 2026
Complete pricing guide with plans, and cost analysis
CoreWeave pricing ranges from $10 to $68.80/instance/hour.
CoreWeave costs $10 to $68.80 per instance/hour as of April 2026, with 4 plans available. Pricing depends on your chosen tier, contract length, and negotiated discounts.
Use the interactive pricing calculator to estimate your exact cost based on team size and requirements.
- Free tier: No free tier available
CoreWeave offers 4 pricing tiers: On-Demand (8x GPU), On-Demand (1x GPU), Spot, Reserved. The On-Demand (1x GPU) plan is inference and smaller workloads.
Compared to other ai/gpu cloud compute software, CoreWeave is positioned at the budget-friendly price point.
How much does CoreWeave cost?
CoreWeave Pricing Overview
CoreWeave has 4 pricing plans ranging from $10 to $68.80/instance/hour. The On-Demand (8x GPU) plan requires contacting sales for a custom quote and is designed for large-scale ai training. The On-Demand (1x GPU) plan requires contacting sales for a custom quote and is designed for inference and smaller workloads. The Spot plan requires contacting sales for a custom quote and is designed for fault-tolerant batch workloads. The Reserved plan requires contacting sales for a custom quote and is designed for predictable long-term workloads.
This pricing was last verified in April 1, 2026.
All CoreWeave Plans & Pricing
| Plan | Monthly | Annual | Best For |
|---|---|---|---|
| On-Demand (8x GPU) | Contact Sales | Contact Sales | Large-scale AI training |
| On-Demand (1x GPU) | Contact Sales | Contact Sales | Inference and smaller workloads |
| Spot availability: Subject to capacity | Contact Sales | Contact Sales | Fault-tolerant batch workloads |
| Reserved | Contact Sales | Contact Sales | Predictable long-term workloads |
View all features by plan
On-Demand (8x GPU)
- HGX B200 (8 GPUs) at $68.80/hr
- HGX H200 (8 GPUs) at $50.44/hr
- HGX H100 (8 GPUs) at $49.24/hr
- A100 (8 GPUs) at $21.60/hr
- L40S (8 GPUs) at $18.00/hr
- L40 (8 GPUs) at $10.00/hr
- No ingress/egress fees
On-Demand (1x GPU)
- GH200 (1 GPU) at $6.50/hr
- RTX PRO 6000 Blackwell at $20.00/hr
- GB200 NVL72 (4 GPUs) at $42.00/hr
Spot
- RTX PRO 6000 Blackwell at $9.24/hr (spot)
- Preemptible instances
- Up to 54% savings
Reserved
- Up to 60% discount over on-demand
- Committed usage terms
- Contact sales for pricing
How CoreWeave Pricing Compares
| Software | Starting Price | Top Price |
|---|---|---|
| CoreWeave | $10/instance/hour | $68.8/instance/hour |
| Lambda | $0.69/GPU/hour | $6.99/GPU/hour |
Detailed pricing comparisons:
CoreWeave Pricing FAQ
01 How much does CoreWeave cost?
CoreWeave GPU instances range from $10.00/hr (8x L40) to $68.80/hr (8x HGX B200) for on-demand pricing. Popular 8x H100 instances cost $49.24/hr. Spot pricing offers savings up to 54%. Reserved capacity with committed terms saves up to 60% over on-demand.
02 Does CoreWeave charge for data transfer?
No. CoreWeave does not charge for ingress, egress, or data transfer fees. Storage pricing ranges from $0.015/GB/month (cold) to $0.070/GB/month (distributed file storage).
03 What GPUs does CoreWeave offer?
CoreWeave offers NVIDIA GB300 NVL72, GB200 NVL72, HGX B200, RTX PRO 6000 Blackwell, HGX H200, HGX H100, GH200, L40S, L40, and A100 GPUs across North America and Europe regions.
04 What is CoreWeave's spot pricing?
CoreWeave's Spot instances are preemptible and offer up to 54% savings over on-demand rates. For example, the RTX PRO 6000 Blackwell is available at $9.24/hr on spot. Spot is best for fault-tolerant batch workloads that can handle interruptions.
05 CoreWeave vs Lambda: which is better for AI training?
CoreWeave focuses on enterprise-scale multi-GPU clusters and offers more GPU variety including the latest HGX B200 at $68.80/hr for 8 GPUs. Lambda offers simpler access and lower minimum commitments, with H100 SXM at $3.99/GPU/hr in 8-GPU configs. CoreWeave's strength is its enterprise SLAs and no data transfer fees.
Is this pricing incorrect? — we verify and update within 24 hours.