ISP Test
Support lines respond to numbers and timestamps, not "it feels slow." This test is built for that call. It monitors for as long as you choose, checks five independent DNS servers at the same moment, and writes down every problem with when it happened and whether it hit everything at once or just one of them.
Test duration
Short tests are useful for quick checks. Longer tests are better for finding intermittent problems.
Free, no signup. Lightweight — the basic test uses only tiny requests.
Customize: goal, connection type
Connected via (optional — sharpens the diagnosis)
How this test is different from the stability test
It logs, it does not just measure
The stability test gives you a score for a short window. This one keeps a running case file and logs every blip as it happens, and the 2- and 5-minute options give problems that come and go time to show up.
Five destinations at once
Every second it checks OpenDNS, Quad9, Cloudflare, Google and AdGuard together. One provider having a bad moment is not evidence about your connection; all five struggling at the same instant is.
Every incident is logged, not counted
Each drop, lost request or slowdown is written down with its time and how long it lasted, so you can read it out on a call or match it to the provider's own records.
All, some or one
Each incident is tagged with how many destinations it hit. Incidents on all of them are consistent with the connection itself: your line, modem, router or provider. An incident on only one points at that destination or the route to it.
What goes in the report
An evidence summary
How many incidents, how many hit every destination at once, the longest one, and how much time they added up to.
A timestamped incident log
Every incident with the time it started, what kind it was, how long it lasted and which destinations it hit — the kind of detail a technician can check against their own logs.
Latency percentiles, jitter and request loss
Median, average, P95, P99 and maximum, so the report holds more than one number that can be argued with.
Your test method, stated plainly
So anyone reading the report knows exactly how the numbers were produced and what their limits are.
How to read the result
Incidents that hit every destination at the same moment are the ones worth raising. Because the destinations sit in different places, a problem that shows on every one of them is most consistent with something on your side of the internet: your home network, your modem or the line, or your provider.
Incidents on a single destination are usually not your provider's fault, since that service or the route to it can have a bad moment on its own. This is a hint rather than proof: a browser test cannot see inside your provider's network, so use the log as a starting point for the conversation.
Nothing is uploaded anywhere
This report is built and shown entirely in your browser. There is no cloud copy and no link to share — copy the summary, download the CSV or text file, or print it, and send it yourself however you prefer.
Getting the most out of a call with your ISP
Test on Ethernet
Rules out Wi-Fi as the cause before you call, so the conversation can focus on the line.
Test at the time problems happen
A clean test from 2pm does not help if your problem is every evening at 8pm.
Ask for a line test from their end
Your browser test shows what you experienced; their equipment can check the physical line, which this test cannot see.