Jitter Test
Check how steady your connection is. The test times 40 round trips, shows how much they vary, then checks ping and real UDP packet loss — in about ten seconds.
Last updated
Times 40 round trips and measures how much they vary
Jitter is the average change in round-trip time between consecutive samples, measured over a warm HTTPS connection to the nearest Cloudflare server. Packet loss is then measured with real UDP packets through a Cloudflare TURN relay — the transport games and voice chat use.
What is jitter?
Jitter is how much your latency changes from one packet to the next. Pings of 30, 31, 29 and 30 ms show about 1 ms of jitter; pings of 10, 50, 15 and 45 ms show 35 ms, with the same 30 ms average. Games and calls need a steady rhythm, so high jitter means stutter and choppy audio.
That is why a ping number alone can look fine while a match feels wrong. The average hides the swings; jitter is the measurement of the swings.
What is a good jitter? Gaming, voice and video calls
Games and call apps handle jitter differently. A game wants every update the moment it is sent, so it feels even small swings. A call app holds incoming audio in a short jitter buffer and plays it out evenly, which lets it absorb more variation at the cost of a little extra delay.
| Jitter | Games | Voice and video calls |
|---|---|---|
| Under 5 ms | Excellent. Nothing to fix. | Clean audio and video. |
| 5–15 ms | Fine for every genre, including competitive shooters. | Absorbed by the jitter buffer; you will not hear it. |
| 15–30 ms | Uneven movement and inconsistent hit registration in fast games. | Still inside Microsoft's 30 ms pass mark for Teams, but the buffer adds delay. |
| Over 30 ms | Visible stutter and rubber-banding. | Robotic or clipped voice, frozen video frames. |
Why jitter matters more for calls than the number suggests
Microsoft's network connectivity test for Teams treats a connection as healthy for calls when UDP jitter is under 30 ms, latency under 100 ms and packet loss under 1%. The jitter buffer is what makes 30 ms workable: it waits long enough for late packets to catch up. But that wait is added to every word you say, and ITU-T G.114 recommends keeping one-way delay for conversation under 150 ms. A connection with high latency and high jitter uses up that budget fast, which is when people start talking over each other.
What causes jitter
Wi-Fi contention. Every device on the same channel, and every neighbouring network, competes for airtime. Waiting for a turn and retransmitting corrupted frames adds a different delay to each packet. This is the most common source of jitter at home.
Bufferbloat. When someone starts a large upload or download, packets queue up in your router or modem. The queue grows and shrinks with the traffic, so your game packets wait a different amount of time behind it on every trip.
Congestion at your provider. Jitter that appears mainly in the evening, on a wired connection, with nothing else busy at home, usually comes from a busy link in your provider's network.
VPNs, mobile and satellite links. A VPN adds an extra hop that can be congested itself. Mobile networks and Starlink switch cells or satellites during a session, and each handover shows up as a jump in latency.
How to reduce jitter
1. Test on a cable. Plug straight into the router and run the test again. If jitter drops sharply, the fix is on the wireless side: use 5 or 6 GHz instead of 2.4 GHz, move closer to the router, or pick a less crowded channel. The Wi-Fi ping test walks through the comparison.
2. Stop what is filling the line. Pause cloud backups, game downloads, console updates and video uploads, then test again. If that fixes it, the problem is bufferbloat.
3. Turn on smart queue management. Many routers offer SQM, often labelled QoS or "gaming mode", using algorithms such as fq_codel or CAKE. They keep the queue short so your game and call packets are not stuck behind a download. Set the limit slightly below your measured speed.
4. Take the VPN out of the path. Run the test with the VPN off. If jitter improves, route games and calls outside the tunnel.
5. Show your provider the results. If jitter stays high on a cable, with nothing else running, at the same time every day, note the numbers and times and report them. A wired result is what rules out your own equipment.
Jitter, ping and packet loss: which test to run
This page measures all three to your nearest server, with jitter as the main result. If the problem is dropped packets rather than uneven ones, the packet loss test also shows whether the losses come in bursts. To compare latency and jitter across 14 regions, use the latency test or the ping test by location, and for a specific title, the game server ping tests.
How this test works
After one warm-up request, the test times 40 small HTTPS requests to the nearest Cloudflare server over the same open connection and reports the average difference between consecutive round trips. That is the same idea as the interarrival jitter defined in RFC 3550, the protocol behind most voice and video calls, without its smoothing (RFC 3550 updates a running estimate by 1/16 of each new difference). The latency shown is the median of the 40 samples.
It then sends 1,000 UDP packets through a Cloudflare TURN relay and counts how many never arrive — the same packet loss measurement as the packet loss test and the full test on the homepage. Browsers cannot send ICMP, so absolute latency reads a little higher than the ping command; jitter is a difference between samples, so that fixed overhead largely cancels out.
Jitter questions
What is a good jitter for gaming?
Under 5 ms is excellent and under 15 ms is fine for any game, including fast shooters. Between 15 and 30 ms you start to feel it as uneven movement and shots that land inconsistently, even when the average ping looks healthy. Above 30 ms fast-paced games stutter noticeably, and the cause is worth tracking down before blaming the game server.
What jitter is acceptable for Zoom, Teams and other calls?
Call apps absorb jitter with a buffer that holds incoming audio for a moment and plays it out evenly, so they tolerate more than games do. Microsoft's own network test for Teams counts UDP jitter under 30 ms as a pass. Above that the buffer either grows, adding delay to the conversation, or runs dry, which you hear as robotic or clipped voice.
Is jitter the same as ping?
No. Ping (latency) is how long one round trip takes; jitter is how much that time changes from one packet to the next. A connection with a steady 60 ms ping can feel smoother than one that swings between 20 and 80 ms, because games and calls can compensate for a delay they can predict but not for one that keeps changing.
Why is my jitter high on Wi-Fi but fine on a cable?
Wi-Fi is a shared radio channel. Your device has to wait for airtime whenever another device or a neighbouring network is transmitting, and it retransmits frames that were corrupted by interference. Each wait adds a slightly different delay, which is jitter by definition. Ethernet has no contention, so the same line tested on a cable usually shows a much lower number.
Why does my jitter change every time I run the test?
Because jitter is caused by whatever else is happening on the path at that moment: another device starting a download, a phone syncing photos, evening congestion at your provider. One run is a snapshot of a few seconds. Run it three times, and at the time of day when your games or calls actually suffer, before drawing conclusions.
Can jitter cause packet loss?
In practice, yes. A game or call app only waits so long for each packet; one that arrives after its moment has passed is thrown away and has the same effect as a packet that never arrived. High jitter and real packet loss also tend to come from the same place, a queue that is overflowing, which is why this test measures both.