Decode Base64 back to text. URL-safe alphabet and missing padding are accepted; invalid input is reported with position instead of printing garbage.
Bad input is reported instead of turned into garbage: characters outside the alphabet, misplaced padding and impossible lengths all produce a message with the position. If the bytes are not valid UTF-8 — an image, a ZIP or encrypted data — the tool says so and shows the raw bytes in hex.
Base64 decoding is often done sloppily: a tool that does not check its input happily prints nonsense when the string is truncated or contains a typo. This decoder validates instead of guessing. It accepts the standard alphabet and the URL-safe variant from RFC 4648, with or without the trailing padding, and ignores line breaks and spaces — which is what you get when Base64 is copied out of an e-mail, a PEM file or a JSON log line. Characters outside the alphabet, padding in the middle and lengths that cannot be a Base64 sequence are rejected with the offending character and its position. After decoding, the bytes are checked against the UTF-8 rules, so you can tell whether the result really is text or binary data such as an image, a ZIP archive or ciphertext; if the bytes are not valid UTF-8 the tool says so and shows them as hex. Everything runs in the browser and nothing is uploaded.