Base64 Encoder and Decoder

What this Base64 tool does

This tool converts Unicode text to Base64 by using the correct text pipeline: Unicode string to UTF-8 bytes to Base64. It also decodes Base64 text back to bytes, then either interprets those bytes as strict UTF-8 text or shows the decoded bytes as hexadecimal. The reverse pipeline is Base64 to bytes to UTF-8 text. Processing runs locally in your browser.

Base64 is byte-oriented. It can represent text, images, compressed files, keys, tokens or arbitrary binary payloads. This page is primarily a text and byte-inspection tool, so invalid UTF-8 is reported separately from invalid Base64 instead of being silently replaced.

Base64 example

Input text:
Hello, café 😀

Standard Base64:
SGVsbG8sIGNhZsOpIPCfmIA=

Standard vs URL-safe Base64

URL-safe Base64 avoids characters that are awkward in URLs and filenames, but it is not automatically safe for every protocol. You still need to consider the surrounding syntax, separators, quoting rules and whether padding is allowed.

Base64 works with bytes, not characters

é
Unicode code point: U+00E9
UTF-8 bytes: C3 A9
Base64: w6k=

Direct btoa("é")-style handling is unreliable because btoa() expects byte-like character values, not arbitrary Unicode text. This tool uses TextEncoder before encoding, so emoji, combining marks and non-Latin text are converted through UTF-8 bytes first.

Base64 padding

Base64 processes bytes in groups of three and emits four encoded symbols. If the final byte group is short, = padding marks the missing bytes. Valid padding has zero, one or two equals signs at the end. An encoded length of one modulo four is impossible. Some protocols omit padding, so this decoder can infer valid missing padding and report when it did so.

Base64 is not encryption

Base64 only changes representation. It does not hide, protect, sign, encrypt, compress or sanitize the data. Anyone with the encoded value can decode it. Do not use Base64 as a security boundary.

Invalid Base64

Invalid input can come from the wrong alphabet, mixed standard and URL-safe characters, impossible lengths, misplaced padding, too much padding or unexpected punctuation. This tool can ignore only ASCII spaces, tabs, line feeds and carriage returns. It does not silently remove non-breaking spaces, zero-width spaces or other Unicode whitespace because those characters may indicate copied-text corruption.

Binary data vs UTF-8 text

Valid Base64 may decode to a PNG, PDF, ZIP file or any other byte sequence. Those bytes are not necessarily UTF-8. In UTF-8 text mode, invalid bytes cause a clear error. Hex bytes mode shows the decoded bytes such as 48 65 6C 6C 6F without assuming a character encoding.

Data URLs

SGVsbG8= is plain Base64. data:text/plain;base64,SGVsbG8= is a Data URL with a media type and prefix. The decoder rejects prefixes by default. If you intentionally enable prefix stripping, it accepts only a valid Base64 Data URL pattern and reports the removed media type.

How to use the tool

1. Choose Encode or Decode.
2. For Encode, choose Standard or URL-safe Base64 and whether to include padding.
3. For Decode, choose Auto-detect or a specific variant.
4. Choose padding and whitespace handling.
5. Select UTF-8 text or Hex bytes for decoded output.
6. Use Advanced options only when you intentionally need normalization before encoding or Data URL prefix stripping.
7. Review warnings and statistics before copying, downloading or using output as input.

Privacy and processing

This tool runs in your browser. Your input is not submitted to UnicodeNow. The local runner writes output as textarea text and uses Blob URLs only for downloads.

Common uses

• Encoding text for APIs.
• Decoding logged values.
• Converting URL-safe Base64.
• Inspecting JWT segment text without verifying signatures or claims.
• Debugging padding issues.
• Verifying UTF-8 payloads.
• Inspecting decoded bytes.

Frequently asked questions