KB vs KiB, MB vs MiB: the difference explained

Ask how big a megabyte is and you get two correct answers, which is the whole problem. The MB vs MiB confusion is not a mistake anyone made—it is two valid systems wearing similar names. Once you see why both exist, the discrepancy between what a drive claims and what your computer reports stops being mysterious.

Two systems, two bases

Decimal units follow the SI prefixes used everywhere in science: kilo means 1000, mega means a million, giga means a billion. So a kilobyte is 1000 bytes and a megabyte is 1,000,000 bytes. Binary units exist because computer memory is addressed in powers of two and 1024 is the nearest power of two to 1000. A kibibyte is 1024 bytes, a mebibyte is 1024 squared, a gibibyte is 1024 cubed. The names KiB, MiB and GiB were standardised precisely so engineers could say "the 1024 one" without overloading the decimal prefix.

Where each convention is used

Storage manufacturers, internet providers and cloud bandwidth bills all use the decimal units, which is why a drive sold as "1 TB" holds a trillion bytes. Operating systems and programming runtimes tend to compute in binary because memory and buffers line up on powers of two. The catch is that many operating systems still print the decimal-looking abbreviations GB and MB even when they are computing in binary, so the same file appears to have two different sizes depending on which screen you read. The byte size converter lets you switch between the two standards so you can read the right number for the context.

Why the gap matters

At small scales the difference is easy to wave away—about 2.4% at the kilobyte level. But it compounds at every step, reaching roughly 5% at mega, 7% at giga and around 10% by the terabyte level. That is large enough to matter in a storage contract, a capacity forecast or a service-level agreement. A monitoring system that ingests raw byte counts but renders them with a binary divisor will visibly disagree with a billing system that uses a decimal one. The fix is to pick a convention per context, label it explicitly then convert at the boundary rather than letting two systems meet silently inside one chart. The next article, why your hard drive shows less space than advertised, walks through the most common real-world version of this.

A mental shortcut

If you only remember one thing: decimal is the marketing and networking number, binary is the operating-system and memory number. When a value looks slightly smaller than you expected, you are almost certainly reading a binary measurement of something quoted in decimal. The number base converter makes the powers of two concrete if you want to see where 1024 comes from.

Frequently asked questions

What is the difference between MB and MiB?
MB is 1,000,000 bytes (decimal) while MiB is 1,048,576 bytes (binary, 1024 squared). The roughly 4.9% gap grows at every larger unit, so the labels are not interchangeable.

Which unit should I use?
Decimal units like MB and GB for storage, bandwidth and customer-facing figures. Binary units like MiB and GiB for memory, buffers and file sizes reported by operating systems.

Bottom line

KB and KiB are not the same and neither are MB and MiB. Decimal units count in thousands, binary units count in 1024s and the gap grows with scale. Keep the two straight with the byte size converter and a confusing few-percent discrepancy becomes a number you can explain.