How to use this tool
- Paste plain text to encode or a Base64 string to decode.
- Choose the operation and standard or URL-safe alphabet.
- Process; copy the text or download the actual decoded bytes.
How the operation works
Base64 maps groups of bytes to an ASCII alphabet. Standard mode requires complete groups with correct padding. Base64URL uses hyphens and underscores and can omit padding. Decoding checks canonical padding bits. UTF-8 decoding is strict: invalid byte sequences produce a hex preview and binary download, not replacement characters.
What Base64 encoding represents
Base64 represents bytes using a limited alphabet. It is useful when a data field can carry text but the underlying value is byte-oriented. Encoding does not hide the value: anyone with the encoded string can decode it without a password. It also does not compress text or prove that a message came from a trusted sender.
This tool first encodes input text as UTF-8. Multilingual text and emoji therefore use their UTF-8 byte sequences, not a one-byte-per-character assumption. The decoder reverses the Base64 representation and then tries strict UTF-8 text decoding. Unicode escapes and code points are different representations, useful when you want to inspect characters rather than encoded bytes.
Base64 encode and decode examples
Hello → SGVsbG8= é → w6k= SGVsbG8= → Hello
The final equals sign is padding for an incomplete group of bytes. It is part of standard padded Base64, not a literal character in the decoded word. A line break at the end of the original text is also a byte and changes the encoding. Check exact source whitespace before comparing two encodings.
For Base64URL, the alphabet uses minus and underscore in place of plus and slash. URL-safe encoding here omits trailing padding. Choose the URL-safe option when decoding that variant; do not manually mix alphabets. A complete JWT contains multiple Base64URL segments, so use JWT Decoder to inspect the token structure.
Invalid Base64 and binary output
Standard decoding requires complete groups and correct padding. Invalid alphabet characters, misplaced padding and noncanonical padding bits stop processing. Surrounding whitespace is trimmed, but line-wrapped content inside the encoded string is not silently removed. Remove only wrapping that you know was added during transport.
Successfully decoded bytes are not necessarily text. If strict UTF-8 decoding fails, the result shows a hexadecimal preview of up to 64 bytes and Download exports the full original bytes as decoded.bin. Copy copies the preview, not an imagined text interpretation. Do not rename arbitrary decoded bytes as an image or executable and assume the content is safe.
Choose the right decoding layer
Use URL decoding for percent escapes and HTML entity decoding for ampersand character references. These operations are not interchangeable. Decode one known layer at a time and inspect the result. For a decoded JSON body, validate and format JSON next.
The current interface processes pasted text, up to 1 MiB per input. Empty input is rejected. It does not fetch URLs, parse data-URI prefixes or encode uploaded binary files. Reset clears the form; it does not erase clipboard or downloaded copies.
A worked example
Hello
SGVsbG8=
Base64 is not encryption and offers no secrecy. A token encoded this way can still contain credentials. Internal whitespace is not automatically ignored; paste a clean encoded string.
Processing and privacy
Base64 encodes or decodes in the page; the text or bytes are not sent to an encoding API. Input stays in browser memory until you reset or leave the page. Copy and download run only when you choose them. No tool input is sent to a processing service. Website pages and scripts still make ordinary hosting requests; the privacy policy separates those from tool processing.
Frequently asked questions
Why does a decoded file show hexadecimal?
Its bytes are not valid UTF-8 text. The binary download preserves them exactly.
Is Base64 encryption?
No. Anyone with the encoded string can decode it without a password. It is representation, not secrecy.
Why does padding matter?
Standard Base64 requires complete groups. Invalid or noncanonical padding stops decoding rather than guessing.
What if the decoded bytes are not text?
The tool shows a hex preview and offers a binary download of the exact bytes.
Base64 or Base64URL?
URL-safe mode uses hyphen/underscore and can omit padding. Match the alphabet to the source system.
How this tool was checked
Base64 was checked with UTF-8 text, emoji, Base64URL, missing or incorrect padding, invalid canonical bits, binary output and empty input. See the testing method or report a failure.