Capacity Block - Pricing Model Pricing Comparison
Capacity Block - Pricing Model
Reserve GPU capacity in blocks. This page provides comprehensive pricing comparison across all major regions, including on-demand, reserved, and spot pricing options. This page explains how capacity block works and when to use it.
Pricing Comparison
| Provider | Pricing Model | Hourly | Savings vs On-Demand | Commitment |
|---|---|---|---|---|
| AWS | Capacity Block | $0.0580 | 40% | 1 year |
| Azure | Capacity Block | $0.0570 | 41% | 1 year |
| GCP | Capacity Block | $0.0560 | 42% | 1 year |
| Alibaba | Capacity Block | $0.0450 | 40% | 1 year |
How It Works
Capacity Block works and when to use it:
- Commitment: You commit to a specific usage level or time period
- Discount: In exchange, you receive a discount off the on-demand price
- Flexibility: Some models allow you to change instance types or regions
- Billing: The commitment is billed regardless of actual usage
Recommendation
For capacity block:
- Analyze usage patterns: Use historical data to determine the right commitment level
- Start small: Begin with a conservative commitment and increase over time
- Monitor utilization: Track reserved capacity utilization to avoid waste
- Combine models: Use a mix of on-demand, reserved, and spot for optimal cost
- Review regularly: Cloud pricing changes; review commitments quarterly
Data Source
- API: Multi-provider Official Pricing API
- Fetched: 2026-07-23
- Methodology: Prices retrieved via official API and normalized to USD hourly/monthly rates