Calculator
Tip: start with realistic requests, payload size, and cache hit rate. Then tune compute and delivery options to match performance goals.
Example data table
These examples show typical input patterns and outcomes using default rates.
| Scenario | Requests/month | Payload (MB) | Cache hit | Egress (GB) | Compute sizing | Security | Reliability |
|---|---|---|---|---|---|---|---|
| Starter site Blog + light API |
800,000 | 1.0 | 75% | 180 | 2 vCPU, 4 GB RAM | WAF: Off | Multi-zone: Off |
| Business app Dashboards + uploads |
4,500,000 | 1.8 | 65% | 900 | 4 vCPU, 8 GB RAM | WAF: On | Multi-zone: On |
| High-traffic store Spiky campaigns |
18,000,000 | 2.2 | 80% | 3,600 | 8 vCPU, 16 GB RAM | WAF + DDoS | Multi-zone: On |
Formula used
- Performance multiplier = Peak factor × (1 + Headroom%/100)
- Compute cost = (vCPU × Hours × vCPU_rate + RAM × Hours × RAM_rate) × Multiplier
- Storage cost = Storage_GB × Storage_rate
- IOPS cost = IOPS × Hours × IOPS_rate
- Egress cost = Egress_GB × Egress_rate
- Request cost = (Requests/1,000,000) × Requests_rate
- CDN cost = Egress_GB × CDN_egress_rate + (Requests/1M) × CDN_requests_rate
- Backup cost = Backup_GB × Backup_rate × (1 + 0.75 × Retention_days/30)
- Support cost = Support_flat + Support_% × (Compute + Storage + Egress + CDN)
- Subtotal = Sum of all components
- Pre-tax = Subtotal × Region_multiplier × SLA_multiplier × (1 − Discount%)
- Total = Pre-tax × (1 + Tax%/100)
All unit prices are editable. Replace them with your provider’s published rates for a precise bill estimate.
How to use this calculator
- Enter monthly requests, payload size, and cache hit rate to describe your traffic pattern.
- Add egress (GB) from analytics or your current bill. Adjust peak factor and headroom to match performance goals.
- Pick a compute billing model. Use instance sizing for fixed servers, or usage inputs for serverless-style billing.
- Set storage, backups, and IOPS based on your database and file needs. Faster disks generally cost more.
- Toggle security, monitoring, and multi-zone reliability. These trade cost for risk reduction and stability.
- Apply region, SLA, discounts, and taxes. Then submit to view a full breakdown above the form.
- Use the export buttons to save reports for budgeting.
Cost drivers that dominate monthly spend
For most workloads, compute, egress, and storage create the largest baseline. A 2 vCPU and 4 GB instance running 730 hours can cost less than 600 GB of outbound transfer, depending on your egress rate. Use the breakdown to see which line items exceed 20% of subtotal, then optimize that layer first.
Sizing for peaks without overspending
The calculator applies a performance multiplier: peak factor × (1 + headroom). If your average load needs 2 vCPU but campaigns spike 1.8× and you keep 25% headroom, the multiplier becomes 2.25×. That means fixed capacity pricing behaves like paying for 4.5 vCPU on average.
Traffic economics: payload, cache, and CDN
Estimated traffic from requests is computed as requests × payload ÷ 1024 to approximate GB. At 4.5 million requests and 1.8 MB average payload, the implied transfer is about 7,910 GB. With a 65% cache hit rate, origin traffic drops to roughly 2,770 GB and 1.58 million origin requests. Cutting payload by 30% or raising cache hits by 10 points can save hundreds of GB of origin transfer and reduce compute needed for peak delivery. Enabling a CDN can shift cost from origin compute to delivery fees, but it usually improves latency and stabilizes peaks.
Storage, IOPS, and backup retention planning
Disk type affects both price and responsiveness. HDD is typically lowest cost for archives, SSD suits general production, and NVMe helps low-latency databases. Provisioned IOPS can raise the bill quickly because it is charged per hour. Backup cost scales with retention using a factor that increases with days retained; moving from 14 to 30 days can raise backup charges by roughly 40% using the model. Consider lifecycle policies and cold storage for long retention.
Governance: region, availability, discounts, and tax
After the subtotal, region and SLA multipliers model higher-cost locations and stricter availability targets. A 1.10 region factor and 1.12 SLA factor together raise cost by 23.2% before discounts. Commitment discounts reduce the result, so apply them only when workloads are predictable. Run three scenarios—baseline, peak season, and incident mode—to test budgets. Finally, enter tax to align budget numbers with finance reporting and procurement approvals.
FAQs
How do I choose between instance monthly and usage based billing?
Use instance monthly when you run steady servers most hours. Use usage based when autoscaling or serverless varies by traffic. Compare both to see which better matches your real utilization and peak behavior.
What should I enter for peak factor and headroom?
Peak factor is your busiest-hour load divided by average. Headroom is extra capacity to keep response times stable. Start with 1.3–1.6 peaks and 15–30% headroom, then refine using monitoring data.
Why does the calculator estimate egress from requests and payload?
It provides a cross-check for your egress input. If your bill shows much lower egress, you may have smaller payloads, higher compression, more caching, or more internal traffic than user-facing traffic.
How do I use custom rates accurately?
Enable custom rates and paste unit prices from your provider quote or invoice. Keep all rates in USD, then set the currency exchange rate for reporting. This keeps the model consistent and the exports easy to audit.
Does enabling a CDN always reduce total cost?
Not always. CDN fees add delivery charges, but they can reduce origin compute and smooth peaks. If cache hits are high and egress is large, CDNs often help. Test with the toggle to compare outcomes.
What is included in the PDF and CSV exports?
Exports include your inputs, the USD cost breakdown, and totals such as pre-tax and grand total. Use CSV for budgeting spreadsheets and PDF for approvals. Recalculate after changes so downloads reflect the latest scenario.