How to Test Proxy Speed: Latency, Throughput, Success Rate

"Fast" can mean three different things for a proxy, and a provider's headline number rarely tells you which. Measure latency, time to first byte, download speed and success rate, and measure them the same way for every proxy you compare.

Which four numbers matter when testing proxy speed?

• Connect time. How long it takes to open a connection to the proxy itself. It reflects the distance and path between you and the proxy.
• Time to first byte (TTFB). How long until the first part of the answer arrives through the proxy, including the proxy's own work and the website's. This is the number that makes pages feel fast or slow.
• Throughput. How many megabits per second you can download through it. It matters for large files and media.
• Success rate. The share of requests that complete without an error or a block. A fast proxy that fails one request in five is slower in practice than a steady one.

Test with the proxy checker

Enter the proxy in the proxy checker. It reports connect time, time to first byte through the tunnel, the download rate of a 1 MB sample and the success rate over three requests. Use it to compare proxies against each other and to catch ones that are broken or very slow. Its limits are listed on the page: it runs from a Cloudflare data center near you, not from your own server, and one sample over one connection reads lower than a long download.

Test from the machine that will use the proxy

The speed you get depends on the path from your machine, so confirm with cURL there. This command prints three timings:

curl -x http://user:password@host:port -o /dev/null -s -w "connect=%{time_connect} ttfb=%{time_starttransfer} total=%{time_total}\n" https://www.cloudflare.com/cdn-cgi/trace

To measure throughput, download a larger file through the proxy and divide its size by the time taken. Run each test several times and at different times of day, and ignore the first result, which includes setup costs.

What to compare

As a rough guide, datacenter proxies usually answer with a TTFB in the tens to low hundreds of milliseconds, while residential and mobile proxies are typically several times slower and less consistent, because the exit is a consumer connection. These ranges vary by provider and distance, so the useful comparison is between your own candidates, measured identically. For background on the difference, read residential vs datacenter proxies.

Why is my proxy slow?

• Distance. A proxy far from you adds latency to every request. Pick a location near you or near your target, depending on which matters more.
• Your own connection. Check your baseline with the speed test and the ping test, so you do not blame the proxy for a slow link.
• Overloaded pools. Shared datacenter IPs and busy residential pools slow down at peak hours.
• The target. Some websites throttle or delay traffic from proxy ranges. Test the same proxy against a neutral page and against your real site to see which one is slow.
• Protocol overhead. It is small. Switching between HTTP and SOCKS5 rarely changes the result.

A simple buying checklist

1. Ask for a trial or a small plan.
2. Test each candidate in the checker, then from your own machine.
3. Record TTFB, throughput and success rate at two different times of day.
4. Test your real target site, not just a test page.
5. Buy the one that wins on your numbers, not the one with the best headline figure.

If a test fails outright, see proxy errors and fixes. For the overall method, read how to check if a proxy is working.