Copy storage pricing: what you pay for when data moves

"Copying data" in cloud storage can mean many things: an intra-region copy, cross-region replication, backup copy traffic, or syncing objects between buckets. The costs usually fall into a few predictable buckets. This guide gives you a checklist so you don't miss a line item.

1) Transfer / data movement

If your copy moves data across regions (or out of a provider), you may pay a per-GB transfer fee. Sometimes a replication feature has its own per-GB charge, sometimes it's billed as cross-region transfer, and sometimes it's both.

Estimate the GB/month that actually changes and gets copied. For most workloads, copy volume is driven by writes and churn, not the total dataset size.

Tools: Storage replication cost and Cross-region transfer cost.

If you're specifically pricing S3 replication, start with S3 replication cost and treat replicated GB/month as the key input.

2) Extra storage at the destination

A copy or replica usually means you're storing the data twice. Destination storage is often the largest hidden cost because it's continuous (GB-month), even if the copy job runs once.

Tool: Object storage cost (add the replica stored GB as another line item).

3) Request fees (PUT/GET/LIST) can matter

Copy and replication workflows can increase request volume (PUT/COPY, GET, LIST, HEAD). For large objects with few operations, requests are often negligible; for millions of small objects, requests can become material.

Tools: S3 request cost and object storage cost.

4) Egress (when copying feeds another system)

If copy is part of a pipeline (for example, copy from storage into analytics, or into another cloud), egress pricing may dominate. Model it explicitly if you expect significant downloads.

Tool: Data egress cost.

5) A quick way to estimate copied GB/month

If you don't know copy volume, start from write throughput (GB/day) or batch sizes (GB/job). Multiply by days/jobs per month. For change-heavy datasets, copy volume is usually closer to writes than total stored GB.

Quick checklist

  • How many GB/month are copied (changed data), not total stored GB?
  • Is there a per-GB replication feature fee?
  • Is there cross-region transfer pricing in addition to feature fees?
  • How much extra storage is held at the destination (and for how long)?
  • Are request fees material (many small objects)?
  • Is there egress out of storage to another service/provider?

Related tools

Storage pricing calculator Replication cost Cross-region transfer Object storage cost


Related guides


Related calculators


FAQ

What is the biggest cost driver when copying data?
Usually the destination storage you keep (GB-month). For cross-region copies, transfer-like charges can also be material.
Should I model based on total stored data?
Usually no. Ongoing replication/copy volume is typically driven by changed data (writes and churn), not the entire dataset size, unless you're doing a one-time backfill.
Do request fees matter?
They can for millions of small objects or workloads with heavy LIST/HEAD usage. For large objects with few operations, request fees are often negligible.

Last updated: 2026-01-21