How it works
- 01
Paste the Base64
Line breaks, missing padding and URL-safe characters are all fine as they are.
- 02
Read the output
Text appears in the second pane. Bytes that are not text offer a download instead.
- 03
Copy or save
Copy the text, or save the bytes as a file with the right extension.
What the decoder tolerates
Real Base64 rarely arrives clean. Base64 copied out of an email carries a line break every 76 characters. JWT segments have their padding stripped. URL-safe variants use - and _ in place of + and /. All three are accepted here without you saying so. Whitespace is ignored, the padding is worked out from the length, and the alphabet is read from the characters present.
What cannot be repaired is a string that is cut short. Base64 packs three bytes into four characters. So a length that will not divide up means characters are missing, and the end of the data went with them. The tool reports that case instead of decoding it into rubbish.
- Accepts
- Both alphabets, padded or not, with any whitespace.
- Output
- Text in your chosen encoding, or a file download.
- Detection
- Spots JSON, images and common file types in the output.
- Errors
- Bad characters and impossible lengths get a position.
Decoding binary
Not everything decodes to readable text. If the bytes are a PNG, a PDF or a zip, the tool reads the file signature first. It then offers a download with the right extension, instead of printing control characters into the page.
The text check is strict on purpose. Bytes that are not valid UTF-8 are reported as not text. Showing them with replacement characters scattered through would look like a decoding bug, when the real problem is the wrong format.