How this jitter test works
Your browser sends 34 tiny HTTP requests to this site and times each round-trip. The first 4 are warm-up: they absorb the TCP and TLS connection setup and are discarded so they do not fake a jitter spike. From the remaining 30 samples, jitter is the mean absolute difference between consecutive round-trips — the same formula (RFC 3550) VoIP systems use for their jitter buffer. Everything is computed locally; nothing about you is recorded.
Method note: browser tests measure HTTP round-trips, so absolute values sit a few milliseconds above a console ICMP ping. The jitter spread is what matters, and it is not affected by the transport.
What is a good jitter value?
| Jitter | Reading | Where it matters |
|---|---|---|
| < 5 ms | Excellent | Competitive gaming, 4K calls, pro VoIP — no issues |
| 5–15 ms | Good | Casual gaming, HD calls, streaming — most users notice nothing |
| 15–30 ms | Fair | Browsing fine; competitive players see micro-stutters |
| > 30 ms | Poor | Robotic audio, rubber-banding, video freezes |
Streaming tolerates jitter by buffering seconds of video; real-time apps cannot. Judge your line by calls and games, not by whether videos play.
Jitter vs ping vs packet loss
| Metric | Question it answers | Test |
|---|---|---|
| Ping (latency) | How long does one round-trip take? | Ping test |
| Jitter | How much does that time vary packet to packet? | This page |
| Packet loss | How many packets never arrive at all? | Packet loss test |
| Bufferbloat | How much worse does latency get under load? | Bufferbloat test |
All four come from the same probing engine and often share one root cause. The classic combination — low idle ping, high jitter that appears only during downloads — is bufferbloat, and the fix is SQM/CAKE on your router.
Why your jitter is high
- Wi-Fi — the number one cause. Every wireless hop adds variable airtime waits; 2.4 GHz with neighbors on the same channel is jitter soup. Test wired before blaming anything else.
- Bufferbloat — queues fill during uploads/downloads and drain unpredictably. If jitter is calm when idle and wild when loaded, that is your router, not your ISP.
- Router CPU — a weak CPU pushed by VPN encryption or heavy NAT varies packet-to-packet handling time.
- ISP congestion — evening peaks on oversubscribed segments swing latency. If it correlates with 8–11 pm, gather evidence and escalate; see how different ISPs compare for gaming.
Related tools
See how this method compares to Waveform, Speedtest.net and Cloudflare's speed test, or go straight to the SQM settings that fix jitter at the source.
FAQ
What is jitter?
Jitter is the variation in round-trip time between consecutive packets. Ping tells you the average delay; jitter tells you how unpredictable that delay is. Real-time applications — VoIP calls, video meetings, online games — suffer from unpredictability more than from a slightly higher steady ping, because their jitter buffers are tuned for a steady rhythm. Jitter is measured in milliseconds, typically as the mean absolute difference between consecutive samples (the RFC 3550 method this test uses).
What is a good jitter value?
Under 5 ms is excellent — good enough for competitive gaming, 4K video calls and professional VoIP. 5–15 ms is good for casual gaming, HD calls and streaming. 15–30 ms is fair: standard browsing is fine but competitive players notice micro-stutters. Above 30 ms is poor — VoIP breaks up, games rubber-band, video freezes even when the average ping looks fine.
What is the difference between ping and jitter?
Ping (latency) is how long one packet takes to travel there and back. Jitter is how much that time varies between consecutive packets. A 30 ms ping with 2 ms jitter feels far better on a call than a 20 ms ping with 40 ms jitter: applications suffer unpredictability more than a slightly higher steady delay. Jitter is computed as the mean absolute difference between consecutive samples — the RFC 3550 method VoIP systems use.
Why is my jitter high but my ping is low?
Low average with wild swings usually means Wi-Fi interference (2.4 GHz is the worst), other devices bursting traffic on the link, or bufferbloat: latency that only appears when the line is loaded. Test on a wired connection to rule out Wi-Fi, then run a bufferbloat test — if jitter explodes only while the line is saturated, SQM/CAKE queue management on your router is the fix.
Does a VPN increase jitter?
Usually yes. A VPN adds an extra network leg and encryption step, and on a CPU-limited router that processing time varies packet to packet — which is jitter by definition. WireGuard typically adds less variation than OpenVPN. Measure with and without to see your real cost.