Byte Size Formatter (KB / KiB / MB / MiB)

Why does "1 KB" mean 1000 bytes in some places and 1024 in others? The short answer: historical confusion. Decimal (SI: 1 KB = 1000) is what disk manufacturers and IEEE use; binary (IEC: 1 KiB = 1024) is what operating systems and RAM use. This tool converts in both directions: enter a byte count and see SI + IEC representations; or parse a human string like "1.5 GiB" back into exact bytes.

How to use the Byte Size Formatter (KB / KiB / MB / MiB)

Mode Bytes → human: enter a raw byte count, get formatted strings in every unit. Mode Human → bytes: enter a human string (1.5 GiB, 100MB, 4096) and get the exact byte count.

The parser accepts both SI (KB, MB, GB, TB) and IEC (KiB, MiB, GiB, TiB) units. Bare numbers are treated as bytes. Optional spaces and case-insensitive matching.

About Byte Size Formatter (KB / KiB / MB / MiB)

The byte-size confusion has a long history. Through the 1980s and 90s, "1 KB" colloquially meant 1024 bytes — the convenient power-of-2 close to 1000. This made sense because memory chips, addressing, and block sizes were all powers of 2. Disk manufacturers, wanting bigger-looking numbers on their packaging, started using the strict SI definition (1 KB = 1000 bytes), which made a "500 GB" drive show as "465 GB" in your OS. Lawsuits ensued (Western Digital settled one in 2007).

The IEC stepped in with new prefixes in 1998: kibi-, mebi-, gibi-, tebi- for the binary powers. So KiB = 1024 bytes, MiB = 1024² bytes, etc. The OS world has slowly adopted these: macOS Finder shows file sizes in SI (1 MB = 1,000,000 bytes); Windows File Explorer shows them in binary but labels them with the SI symbols (showing "1 MB" but meaning 1,048,576 bytes); Linux ls -h uses powers of 1024 with SI labels by default, IEC labels with --si.

For network and storage planning, knowing which scheme a tool uses matters: a "10 MB/s" network link transfers 10,000,000 bits/sec (decimal); a "1 TB" disk holds ~931 GiB (a 7% gap that adds up at scale). When in doubt, this tool shows both side-by-side.

Common use cases

• Capacity planning — "I need to store 500 million 50-KB documents" — quick total in human units.
• Bandwidth math — "downloading 50 GiB at 100 Mbps takes how long?"
• API design — picking an upload limit ("max 10 MiB per file") and expressing it in bytes.
• Reading docs — when a config says "memory_limit: 512M", which definition does it use? (Hint: in most stacks it's 512 × 1048576 = 537 million bytes.)
• Cost estimation — converting cloud storage / egress prices ($0.023 per GB-month) into your scale.