Input
Mode:
Output
Formatted result will be displayed here
Base64 decoding walks the input four characters at a time, converts each to six bits, and reassembles them into the original bytes. This decoder is tuned for developer quick-checks: paste a value with or without padding, with or without line breaks, and it returns the UTF-8 text without fuss. If the bytes do not form valid UTF-8, it tells you so you can switch to a binary-aware tool.
Secret values printed as Base64 in CI logs decode in seconds so you can confirm what actually shipped to the runner.
An inbound webhook carrying Base64 becomes readable immediately, making integration debugging much faster.
Sometimes data gets double-encoded by mistake. Decoding once or twice confirms which layer is redundant.
When a blob looks like it might be Base64, decoding is the fastest test. Gibberish back means it was not.
The tool tolerates missing padding. Many real-world Base64 values, especially in JSON, drop the trailing equals signs.
Yes. Line breaks are ignored so MIME-style wrapped Base64 decodes exactly the same as a single-line string.
Use the Base64URL decoder for that variant. Characters - and _ are not accepted by the standard decoder.
No. Decoding happens inside your browser tab, and nothing is sent to any server.