Cloud & Platform

Data Egress Cost Calculator

The same bytes costed to six destinations, with roughly a factor of nine between the cheapest and the dearest.

Loading the tool…

Processing happens locally in your browser. What you paste or load is processed by this page and is not uploaded to a server. Nothing is stored unless you use a control that says it stores something, and you can clear anything this site has kept from the privacy page.

How to use this tool

  1. Set the volume that leaves per transfer, and how many transfers happen in a month.
  2. Set the free allowance your account gets — 100 GB a month is the common figure.
  3. Select Cost the transfer.
  4. Find the row that matches where the data actually goes, then check whether it could go somewhere cheaper — or not at all.

What data egress cost calculator does

Data in is free on every major cloud. Data out is not, and the price depends entirely on where it is going: free within an availability zone, a cent a gigabyte between them, two cents between regions, and nine cents to the internet. Almost every shocking egress bill is a workload somebody assumed was in the first category and was actually in the last.

This costs the same volume to all six destinations at once, monthly and annually, so the gap is visible rather than theoretical. Note the cross-zone line in particular: AWS charges it in both directions, so a chatty service split across zones pays twice for every round trip. And before you negotiate a rate, look at the volume — the cheapest egress is the transfer that does not happen, and compressing it or moving the compute to the data usually beats any discount.

Frequently asked questions

No. Ingress is free on all three major providers, which is deliberate: getting data in is easy and getting it out is not. That asymmetry is the thing to keep in mind when a design casually moves large volumes back and forth across a boundary.

On AWS, traffic between availability zones is billed on both the sending and the receiving side, so a round trip between two services in different zones pays for both directions. It is a small per-gigabyte rate that becomes significant for a chatty service, and it is the reason latency-sensitive components are usually pinned to one zone with redundancy handled at a higher level.

The per-gigabyte rate is similar, so the saving does not come from the price. It comes from the cache: the origin pays egress once per object per cache period rather than once per request, so for content requested many times the origin bill falls sharply. For data requested once, a CDN adds a hop and saves nothing.

In order: not transferring the data, which usually means moving the compute to it; transferring less of it, through column pruning, aggregation before export or a more efficient format; and only then negotiating the rate. Committed-use discounts and CDN deals are real but they are the last lever, not the first.