Watch it happen: decoding “hello”

Take 01101000 01100101 01101100 01101100 01101111. Split into bytes, and each byte is a number: 104, 101, 108, 108, 111. Look those up in the binary letters chart and you get h, e, l, l, o — “hello”. That is everything decoding ever does; the converter above just does it instantly and catches the mistakes along the way.

Three ways to decode binary

  • With this binary to text tool — paste anything and it auto-detects the direction, cleans messy formatting (spaces, commas, 0b prefixes), and flags broken bytes instead of guessing.
  • By hand — split into bytes, add the place values, read the numbers off the ASCII table. Our guide to decoding binary by hand walks through it in four steps.
  • With one line of code — in JavaScript, new TextDecoder().decode(new Uint8Array(bytes)) turns an array of byte values back into text.

Common binary decoding errors — and how to fix them here

  • Wrong bit count — text bytes are exactly 8 bits, so a 7-bit group means a bit was lost in copying. The Error Clinic names the exact group and offers to pad it with a leading zero or drop it — and when the whole paste divides evenly by 7 (old-school 7-bit ASCII), it also offers a one-click Regroup as 7-bit codes.
  • Wrong encoding — if the output looks like gibberish or stops with an error above value 127, the data is probably UTF-8 (Chinese, emoji, accented letters) while the encoding is set to ASCII. Switch it under Advanced settings.
  • A number mistaken for text — 10101 is not a broken letter; it is the number 21. The decoder asks before interpreting ambiguous input instead of inventing a character. More on that in binary numbers vs binary text.

One honest limit: decoding tells you what the bytes say — it cannot tell you whether they were ever meant as text. Binary from an image or a file will “decode” into nonsense, and no tool can prove otherwise. When in doubt, check the source.