Binary to Text Converter

How binary becomes text

Computers store text as numbers. Each character is assigned a number by a character set, the number is stored in one or more bytes, and a byte is eight binary digits. Converting binary to text is that process run backwards: cut the digits into groups of eight, read each group as a number from 0 to 255, and look up the character.

01001000  01101001
   72       105        ← each byte as a decimal number
   48        69        ← the same bytes in hexadecimal
   H         i         ← the characters those numbers stand for

Binary, hexadecimal and decimal are three ways of writing the same byte. Binary is what the machine holds. Hexadecimal is what people use to read it, because one hex digit is exactly four bits, so every byte is exactly two digits. Decimal is what you get when a program prints a byte array.

Two patterns in that table are worth remembering. Upper- and lower-case letters differ by a single bit, which is why A is 41 and a is 61. And the digit characters are not their own values: the character 0 is byte 48, not byte 0.

ASCII, UTF-8 and why one character can be several bytes

The first 128 numbers, 0 to 127, are ASCII: the English alphabet, digits, punctuation and control codes. Each fits in a single byte whose first bit is 0. Everything else — accented letters, other scripts, symbols, emoji — needs more room, and UTF-8 provides it by using two, three or four bytes for one character. The first byte's leading bits say how many bytes follow, and every following byte begins with 10.

UTF-8 is the encoding of the web and of nearly every API, and plain ASCII text is already valid UTF-8, so this converter always uses it. A tool that reads each byte as a separate character turns é into é. That two-character garble has a name, mojibake, and it means UTF-8 bytes were decoded as a single-byte encoding somewhere along the way.

If the bytes you paste are not valid UTF-8 — a lone continuation byte, a sequence cut short — the converter says so and shows each byte as its Latin-1 character instead, so you can still see what is there. That usually means the data is not text at all, or was written in an older encoding such as Windows-1252.

Input the converter accepts

An unbroken run must divide evenly into bytes: a multiple of eight binary digits, or an even number of hex digits. If it does not, a digit has been lost or added, and the converter stops instead of guessing where. Decimal always needs separators, because 7210 could be 72 and 10 or 7 and 210.

How auto-detect decides

Input containing only 0 and 1 is read as binary. Input containing any of the letters a to f is read as hex. Digits alone are ambiguous — 48 69 is “Hi” in hex and “0E” in decimal — so the rule is: if any value has three digits it is decimal, otherwise hex. When that guess is wrong for your data, set the format yourself. The bar above the result always shows which format was used.

Where this turns up in API work

• Reading a hex dump. Packet captures, xxd output and database BYTEA columns show bytes in hex. Paste the hex column to read the text.
• Finding an invisible character. Convert a suspicious string to hex and a non-breaking space shows up as c2 a0 where an ordinary space would be 20; a byte-order mark is ef bb bf at the very start. Both break JSON parsers, as the JSON validator explains.
• Checking a signature or hash input. HMAC signatures on webhooks are computed over exact bytes. Seeing the bytes shows whether a line ending is 0a or 0d 0a.
• Byte arrays in logs. A language that prints [72, 101, 108, 108, 111] is showing decimal bytes; paste the numbers to read them.

Hex output from a hash function is a different matter: a SHA-256 digest is 32 arbitrary bytes, not encoded text, and decoding it produces noise. The same goes for encrypted or compressed data. When binary data needs to travel inside JSON or a header, it is usually written as Base64, which packs three bytes into four characters instead of hex's two characters per byte. Bytes inside a URL are written as percent-encoded hex, which the URL decoder handles.

Frequently asked questions

Related tools and reference