CDN cost comparison: how to compare pricing across providers

A reliable CDN cost comparison is not about a single $/GB number. You need a normalized input set that captures bandwidth, requests, regions, and cache fill. Use this guide as a comparison checklist.

Step 1: Normalize your inputs (the comparison worksheet)

• Bandwidth (GB/month): edge delivery to end users.
• Requests (per month): total billable requests (per 10k or per 1M units).
• Region mix: percent of traffic by region or price tier.
• Origin egress (cache fill): GB pulled from your origin provider.
• Peak month: an incident or launch month for stress-testing.

Step 2: Compare the three cost buckets

• Edge bandwidth: apply $/GB for each region mix and sum.
• Request fees: convert request units (per 10k vs per 1M) to a common unit.
• Origin egress: model separately because it is billed by your origin provider, not the CDN.

Step 3: Adjust for tiering and discounts

• Tiered pricing: use a blended rate or compute each tier explicitly.
• Commit discounts: apply committed-use discounts to both bandwidth and requests if applicable.
• Free tiers: subtract allowances before applying rates.

Step 4: Validate with two scenarios

• Baseline: typical month with your normal cache hit rate.
• Peak: launches, spikes, or incidents (higher requests and cache miss rate).

A provider that looks cheaper at baseline can become more expensive at peak if request pricing or regional mix is unfavorable.

Step 5: Compare add-ons and hidden line items

• WAF or bot protection: request-based charges can dominate during attacks.
• Logs and analytics: log ingestion and query scans can be a second bill.
• Edge compute: functions at the edge add per-request compute costs.

Step 6: Verify billing units and measurement windows

• Confirm GB vs GiB and per-10k vs per-1M request units.
• Check whether pricing uses regional tiers or global blended rates.
• Note whether free allowances apply per month or per account.

Comparison template (fill with your numbers)

Validation checklist

• Cross-check edge GB and request counts from analytics, not just invoices.
• Validate cache hit rate by path; dynamic endpoints skew comparisons.
• Run a peak scenario and confirm the ordering still holds.

Contract and migration considerations

• Minimum commits: some providers require bandwidth commitments that change effective pricing.
• Overage rates: confirm how overages are priced above your commit.
• Migration cost: DNS cutover, cache warm-up, and validation can create a short-term spike.

Feature parity (cost is not the only variable)

• Compare caching controls, purge behavior, and header normalization.
• Check logging options and retention limits.
• Confirm WAF/bot features and how they are billed.

Performance and reliability checks

• Validate latency and cache hit rate in your top geographies.
• Review POP coverage and regional capacity for peak events.
• Confirm SLA credits and support response times for outages.

Data for a fair comparison

• Use 30 to 90 days of traffic to smooth out weekly cycles.
• Break out bot traffic and synthetic monitoring if possible.
• Include large downloads and API endpoints separately.

Common comparison mistakes

• Comparing only $/GB and ignoring request fees.
• Using a single region rate for a global traffic mix.
• Ignoring origin egress and cache fill.
• Using peak rates for a baseline comparison (or vice versa).

Use calculators to run the comparison

Related guides

Related guides

Related calculators

FAQ