Short continuous observation
Catch stalls that a one-shot test can miss.
Repeated probes show latency drift, visible spikes and HTTP request failures while you reproduce a call, Wi-Fi or VPN problem.
Live measurement
Connection Stability Monitor
Keep this tab visible. Background timer throttling can break the sample interval.
This test uses only small requests, but background tabs, extensions and battery saving can still affect timing.
Results will update live on this page.
This run
Network quality summary
Translate it into real tasks
These are transparent threshold-based guides, not service guarantees.
VPN deltas versus direct baseline
Method
One reachability probe per second
The monitor requests a zero-byte, no-store Cloudflare Speed resource once per second. It reports successful responses, request failures, median latency, jitter and the slowest measured point. Every sample remains inspectable.
A concrete example
Reading a short dropout
If 29 of 30 HTTP probes succeed and one times out, the page reports 96.7% HTTP reachability and one failed request. It does not call that 3.3% packet loss because HTTP and UDP packet loss are different measurements.
Limits and uncertainty
What these numbers can—and cannot—tell you.
- 01
A 15–60 second window can confirm a reproduced stall but cannot prove all-day reliability.
- 02
Browser timer throttling can affect a background tab. Keep this page visible during the observation.
- 03
HTTP failures are an application-path signal, not an estimate of UDP or ICMP packet loss.
Anonymous feedback
Did this tool solve the job?
Found a problem or an idea to make this tool better? Tell us. We review every report and start working on actionable improvements within 2 days, continuously improving reliability and usability.