A speed test gives you a few numbers — and the two people confuse most are Mbps and ping. They measure completely different things, and which one matters depends on what you do online.
Test your speed & ping →Free · No signup · ~10 secondsMbps (megabits per second) is your bandwidth — how much data moves each second. Higher is faster. It decides how quickly files download, how many devices stream at once, and whether 4K loads instantly. Think of it as the width of the pipe.
Ping (measured in milliseconds) is latency — the delay for a small piece of data to travel to a server and back. Lower is better. It decides how responsive your connection feels. Think of it as the reaction time.
You can have huge Mbps but still feel laggy if ping is high. A 1 Gbps connection with 200 ms ping feels worse in a game than a 50 Mbps connection with 15 ms ping. Bandwidth moves big loads; latency makes everything feel instant.
| Metric | Good value |
|---|---|
| Download (Mbps) | 100+ for a household |
| Upload (Mbps) | 10+ for calls |
| Ping (ms) | Under 50 (under 20 = excellent) |
| Jitter (ms) | Under 10 |
Run a quick speed test to see your download, upload and ping together. If your Mbps is fine but ping is high, that's usually a Wi-Fi or routing issue — see how to make your internet faster.
Think of a motorway. Mbps is how many lanes it has. Ping is how long the journey takes. Adding lanes moves more cars per hour; it does not shorten the drive.
That's why a gigabit connection can feel sluggish and a modest one can feel instant. Almost everything that feels like “slow internet” in normal use — a page hesitating before it loads, a call dropping a syllable, a game registering your shot late — is the journey time, not the lane count.
Ping is the average delay. Jitter is how much that delay varies. A connection with a steady 40 ms ping feels better in a call or a game than one bouncing between 10 and 90 ms, even though the second one has a lower best-case number.
Jitter is the reason a call breaks up on a connection whose speed test looks perfect. Audio and video arrive in a rhythm; when the spacing scatters, the app has to either wait (delay) or drop what came late (stutter). Under 20 ms of jitter is comfortable; over 50 ms is audible. We cover it, and its companion problem, in jitter and packet loss.
Here's the one that explains a lot of otherwise baffling complaints. Network equipment has memory buffers to hold packets during a burst. When those buffers are too large — which is extremely common in consumer routers — a heavy download fills them, and every other packet has to queue behind it.
The result: your ping is 15 ms when the line is quiet and 300 ms the moment someone starts a download. The speed test still reads brilliantly, because it's measuring throughput, and throughput is fine. Meanwhile calls break up and games become unplayable whenever anyone else in the house does anything.
You can spot it yourself: check your ping while idle, then start a large download and check again. If it multiplies several times over, that's bufferbloat. The fix is enabling Smart Queue Management or SQM in the router if it offers it — not a faster plan, which makes it slightly worse by filling the buffers faster.
| What you're doing | Watch this |
|---|---|
| Competitive gaming | Ping & jitter |
| Video calls | Jitter & upload |
| 4K streaming | Mbps |
| Large downloads | Mbps |
| General browsing | Ping |
| Cloud backup | Upload Mbps |
Notice how rarely raw download speed is the answer. It's the number that gets advertised because it's the number that's easy to sell, not because it's the one you feel.
Two tests on the same connection can report noticeably different ping, and neither is lying. Most time an HTTP request with a stopwatch, which includes connection setup and server processing on top of the actual network delay.
We read the true TCP round-trip time reported by the server's own network stack instead, and report the lowest of six samples. That reads lower than stopwatch-based tests and is closer to the delay your game or call actually experiences. The full explanation is in how we measure.