Ping Monitor: live latency chart with jitter and packet loss

Terminal ping sends ICMP echo requests at the network layer. Browsers cannot do that for security and privacy reasons. This tool measures the HTTP round-trip time instead: it sends a small GET request and times how long the response takes using the browser performance clock. After the first request warms up the connection, the numbers are usually within a few milliseconds of terminal ping for the same host.

Browsers restrict fetch requests to http and https URLs. You can enter https://8.8.8.8/ and if the host serves an HTTPS response, the timing will work. Plain ICMP packets to an IP address are not possible from a web page. The Cloudflare DNS preset uses https://1.1.1.1/cdn-cgi/trace for exactly this reason.

Jitter is the variation between consecutive ping times. A steady 40 ms feels smooth. A wobbling 20 to 80 ms feels terrible on a voice call or in an online game, even though the average is the same. Voice and video tools care about jitter more than average because it is what causes choppy audio and stutter.

For web browsing, anything under 100 ms feels instant and up to 200 ms is fine. For video calls, under 150 ms with under 30 ms jitter is comfortable. For competitive gaming, players want under 60 ms. Anything over 250 ms feels laggy in interactive apps.

A request that never completes before the browser timeout is counted as a lost packet. This is not the same as ICMP packet loss because the request could also fail on the TLS handshake or on CORS. In practice it is a good proxy: if you see losses on every target at the same moment, your line dropped. If only one target loses packets, that host or its CDN is having issues.

The first request to any target has to do a DNS lookup, a TCP handshake and a TLS handshake. That is why the first ping is often 100 to 500 ms higher than later ones. After the browser caches the connection, subsequent requests reuse it and the timing drops to pure round-trip time.

The game you are playing has its own servers, which may be physically closer to you than the preset targets in this tool. Add the game server region or a nearby cloud region as a custom target to get a closer comparison. Also, some games report ping to their match server, not to their login server, so the numbers naturally differ.

No. There is no backend. Requests go from your browser straight to the target host, and the numbers never leave your device. The CSV export downloads locally through a blob URL. Open your network tab to confirm.

Yes. The chart keeps the most recent 300 samples per target. That is 5 minutes at 1 per second, 25 minutes at 5 per second, or up to 25 minutes at a 5 second interval. Longer runs still show correct running averages, min, max, jitter and loss, because those are computed over everything still in memory.

Yes, in modern mobile browsers. Mobile networks often have higher jitter than Wi-Fi, so this is a useful way to see the difference between Wi-Fi and cellular on the same device. Keep the tab in the foreground for best results: browsers throttle background timers.

Each ping downloads a tiny resource, usually a favicon or a small trace file. On the default 1 second interval, a 5 minute run uses under a megabyte on most targets. Pick the 5 second interval if you are on a metered connection.

A gap means a request failed or took longer than the browser timeout. The tool still records it and counts it toward packet loss. Gaps often line up with router reboots, ISP hiccups or your laptop sleeping.

Time to first byte measures how long between sending a request and receiving the first byte of the response. Round-trip time measures a full round trip and back. For HTTP requests with small responses, they are effectively the same number and this tool times the full request, so the values are RTT-style.

Yes. Keep it open in a browser tab or on a second device. If latency and jitter stay low while your call is bad, the problem is likely the call provider. If both spike together, the problem is your network.

Your connection has a limited bandwidth. When other tabs download data, the ping request has to share the same pipe, so it queues behind larger transfers. That is a real measurement of congestion, not a bug in the tool.

The page loads offline after the first visit, but pinging requires network access by definition. If you lose connectivity, every target will report 100 percent loss until the connection is restored.