Jitter Test

How to use

1. Press Start test. Keep this tab in the foreground while it runs. Everything happens in your browser against Cloudflare's network. No app. No account needed.
2. Let the latency phase run. It pings Cloudflare in a fast burst first, and that spread between pings is your jitter. Then it runs download and upload, checking latency again under load for the bufferbloat grade.
3. Read your jitter. Your jitter shows in milliseconds right next to your ping. A number that barely moves means a solid connection. Run it a couple of times. One busy moment on your network can throw off a single reading.

Why test this

Your download speed looks healthy, but the game still rubber-bands, Discord voice breaks into robot noise, and the video call freezes for half a second at the worst moment. Those stutters usually aren't a bandwidth problem. That's jitter: your latency refuses to hold still, and a plain speed test buries that instability inside one average ping number. A line that looks fast on paper can still choke the moment it matters most. What you want is a low number that stays low. A steady 40 ms ping beats one that sits at 20 and hits 150, because your game and your call both plan around the delay they keep seeing. Run a burst of pings back to back and measure how far apart they land. That's the real test of a stable connection, not just a fast one on a headline speed figure.

What the results mean

Jitter is the headline number here, shown in milliseconds. It's how much each ping's timing drifts from the last one. A small figure means your packets are landing right on schedule. Under 5 ms is tight enough that you'll never feel it. Up to about 15 ms is fine for calls and most games. Past 30 ms is where aim starts to feel inconsistent in a shooter and voice chat garbles. Your ping sits next to it as the baseline round trip, and the two read together: low ping with high jitter is still an unstable line. You also get a bufferbloat grade, the part pure jitter tools skip, which times your latency again while the connection is busy. The reading comes from your browser timing real requests to Cloudflare's edge, so it reflects your Wi-Fi, router, and route, not a lab figure. Run it twice and trust the pattern, not one sample.

FAQ

What is a good jitter for gaming and calls?

Under 5 ms is excellent and invisible in practice. Up to about 15 ms is fine for voice calls and most online games. Between 15 and 30 ms you'll notice the occasional hitch. Past 30 ms, competitive shooters like Valorant or CS2 feel inconsistent and voice chat starts to break up. Lower and steadier is always better than a low average that spikes.

Why is my jitter high?

A crowded Wi-Fi channel, distance from the router, an overloaded connection, or a background upload eating your bandwidth all raise it. Older routers pile on more. They buffer too much traffic under load, and that's the bufferbloat this test also grades. One device waking up mid-run can skew a single reading, so test more than once.

How do I reduce jitter?

Plug into Ethernet if you can, since a wired link removes the wireless interference that causes most of it. Close background uploads and cloud backups while you game or call. Reboot an old router, and if yours supports smart queue management or QoS, turn it on to keep latency low when the line is busy.

Does Wi-Fi cause jitter?

Often, yes. Wireless signals share the air with neighbors, microwaves, and every other device on your network, so timing varies more than it does on a cable. The farther you sit from the router, the worse it gets. Testing on Wi-Fi and again on Ethernet is the fastest way to see how much your wireless link is adding.

Is this jitter test accurate?

It measures real round trips from your browser to Cloudflare's global network, reflecting the path your everyday traffic takes, including Wi-Fi and your router. Call it an estimate, not a lab instrument: it reads latency over HTTP, not raw ICMP ping. Packet loss needs a special server setup and stays out of a browser test. Run it a few times and go by the typical result.