Latency

Latency is the time a single request takes to reach its destination and come back, measured in milliseconds — and every proxy adds some, because your traffic makes an extra hop on the way. It is a delay per request, not a limit on how much data can move.

Latency is not bandwidth

These get confused constantly. Bandwidth is how much can flow at once; latency is how long the first byte takes to arrive. A connection can have plenty of bandwidth and still feel slow if each request waits, and a connection with modest bandwidth can feel snappy if the round trip is short. A page that loads slowly on a proxy is usually a latency story, not a bandwidth one.

Why mobile proxies add more of it

A mobile proxy sends traffic over a cellular radio link and through the carrier's network translation before it reaches the open internet. That path is inherently longer than a server-to-server hop, so mobile addresses are slower per request than a datacenter proxy. It is the cost of the trust the carrier address buys you.

When it actually matters

For manual work — logging in, posting, checking placements, running an account by hand — the extra delay is barely noticeable, because a human is the slow part. For a tool that fires thousands of sequential requests, added latency multiplies across every one of them and decides how long the job takes. If that is your workload, run requests in parallel rather than one after another, and pick the proxy type by whether the target scores your address at all.