Where your traffic goes, and why the meter and your logs disagree

The dashboard shows how much traffic each proxy has used, against its total where it has one. Traffic is counted in both directions, what you send as well as what you receive, which is the main reason your own tooling will not match it to the byte. Residential is coming soon and cannot be ordered yet, so today this applies to ISP and datacenter IPs.

Find the figure

• One proxy: open it from its list under Proxies in the side menu. The Capacity card shows Bandwidth used, against the total where there is one.
• Several at once: the Bandwidth column on the datacenter list shows the same figure per row. The ISP list has no such column; open each proxy.

Dedicated and premium ISP and datacenter IPs have no bandwidth cap. Shared static IPs do have a total, the included allowance plus any Extra bandwidth added when ordering, and the Capacity card warns you when it is nearly spent.

Why the numbers disagree

Your scraper counts what it handed to its HTTP library. The meter counts the traffic that went through the proxy.

• What you send counts as well as what you receive.
• Compression shrinks a response in transit, while your tool may count the unpacked body.
• The meter counts bytes, not successful requests: failed and redirected requests use traffic too.
• A client that opens a fresh connection for every request spends more than one that reuses them.

A small gap is normal. A large gap that stays large deserves a look.

Spend less

1. Ask only for what you need, and send Accept-Encoding: gzip.
2. In headless browsers, block images, fonts and video.
3. Reuse connections where rotation does not matter.
4. Stop retrying a request that keeps failing for the same reason.

If the figure looks wrong

It can be on our side, and we want to know. If usage climbs while your clients are stopped, or stays flat while they run, email support@proxyhive.io with the proxy's ID, the time you checked and what you saw. Leave passwords out.