Mbps vs MB/s

ISPs and an internet speed tester speak megabits per second (Mbps). File downloads often speak megabytes per second (MB/s). There are eight bits in a byte, so the same path looks like different numbers depending on the label. A lot of “my internet is slow” threads are this mix-up wearing a disguise. The connection did not shrink by a factor of eight overnight. The window changed units.

Bits are the small flags networks shuffle. Bytes are groups of eight bits, which is how files and disks are usually counted. Marketing likes bits because the number is larger. Operating systems like bytes because people think in files. Neither camp is lying. They are using two rulers. You need to know which ruler is in the screenshot before you call the provider or buy a new plan.

The conversion

Roughly, MB/s ≈ Mbps ÷ 8. A 160 Mbps result is about 20 MB/s. A “10 MB/s” download is about 80 Mbps. If those two screens disagree by a factor of eight, you are probably mixing units — not failing the plan. SpeedMetric reports Mbps, the same unit as the bill, so you can compare the tester with the product you pay for without doing mental gymnastics every time.

The word “mega” also hides a second trap. Network people usually mean 1,000,000 bits in a megabit. Disk tools sometimes mean 1,048,576 bytes in a mebibyte. For home troubleshooting you can ignore that rounding. The bit-versus-byte factor of eight is the one that actually wrecks comparisons. Write the unit next to every number you keep. “90” without Mbps or MB/s is how arguments start.

Why the file download can still be slower

Your last mile can be 200 Mbps while a particular file server sends at 20. The tester talks to a nearby bandwidth endpoint. A game patch talks to a content network that may be busy, far away, or rate-limited. That is not a broken speed tester. It is a different destination. Browsers, app stores and consoles also show progress in ways that average over time, pause for disk, or include handshake delays. A five-second speed test will not match a forty-minute installer graph, even in the same units.

Overhead takes a cut as well. Encryption, Wi-Fi retries and TCP do not appear on the bill. A wired quiet-hour test that lands a little under the plan is common. A download from a sluggish mirror that lands far under the tester is also common. For what download versus upload mean in the first place, read the metrics explained. Then test internet speed on a quiet path so you know the last mile before you blame a file host.

Bits, bytes, and “good enough”

HD and 4K video are often discussed in Mbps because codecs are rated that way. A stream that needs 15 Mbps download does not need 15 MB/s. Disk tools and many browser download panels use MB. Use one unit per comparison. Then decide with what speed you actually need, not which screenshot looks bigger. A household that can hold two streams and a call is “good” even if the marketing number is larger.

When the converted numbers still disagree with the plan after a fair wired test, you have a real gap. That is the moment to read why a speed test can sit below the advertised rate and to follow the debug path. Until you convert, you do not know whether you are looking at physics, a busy house, or a capital B.

A practical habit: when someone texts you a result, ask for the unit. When you keep notes, write “Mbps” every time. When a download window looks tragic, divide the tester by eight and look again. It is unglamorous arithmetic, and it prevents a surprising number of unnecessary upgrades.

Worked examples you can reuse

Suppose the bill says 200 Mbps and SpeedMetric shows 180 Mbps on Ethernet. Divide 180 by eight and you should expect about 22.5 MB/s from a healthy file server. If the browser shows 22 MB/s, the last mile and the file host agree. If the browser shows 4 MB/s, the file host or the app is the bottleneck, not the plan. If SpeedMetric already shows 40 Mbps on the sofa, convert that first — about 5 MB/s — before you decide the download window is “broken.”

Another common pair: a phone says “10 MB/s” in a store download while the tester says 85 Mbps. Those are the same order of magnitude once you convert. The remaining gap is overhead, Wi-Fi, or a busy shop. It is not proof that the provider stole 75 Mbps. Keep the unit in the sentence every time you talk about the result, including when you write to support. “I saw 10” is how tickets go in circles. If you want a live number in the same unit as the bill, use the internet speed test rather than a file-progress bar.

Common questions