UTF-8 to Base64: Encode Unicode Text and Emoji

How to encode UTF-8 text to Base64

1. Type or paste your text into Input · text. Charset is already set to UTF-8.
2. The Base64 appears in Output · Base64. The Input figure under the tool is the UTF-8 byte count; the counter above the box counts characters the way JavaScript does, so an emoji counts as 2.
3. Tick URL-safe (- _) and No padding (=) for URLs and tokens, or pick a Line wrap for email or PEM.
4. Copy the output or download it as base64.txt.

Base64 to UTF-8 turns the result back into text.

How UTF-8 turns characters into bytes

UTF-8 uses 1 to 4 bytes per code point. ASCII stays at 1 byte, which is why English text encodes the same in UTF-8 and ASCII. Everything else takes more:

Since Base64 adds a third on top of the bytes, a Chinese or Japanese sentence grows roughly fourfold from characters to Base64 characters, and an emoji alone becomes 8 characters. Keep that in mind for fields with a length limit; the size calculator helps.

Emoji sequences and skin tones

What looks like one emoji is often several code points glued together. A skin tone is a separate modifier, and family or profession emoji join several people with an invisible zero-width joiner (U+200D):

Cutting a Base64 string or its decoded bytes in the middle of such a sequence leaves broken characters, so never truncate encoded UTF-8 at an arbitrary byte.

Same text, different bytes: Unicode normalization

Some characters have two valid spellings in Unicode. é can be the single code point U+00E9 (composed, NFC) or e followed by U+0301 COMBINING ACUTE ACCENT (decomposed, NFD). Both render identically, but the bytes differ:

File names on Apple’s older HFS+ file system are stored decomposed, while most keyboards produce composed text. If you compare Base64 values, sign them or use them as keys, normalize first. This tool encodes exactly the code points you paste and does not normalize.

UTF-8 to Base64 in code

In JavaScript, TextEncoder always produces UTF-8. Normalize first when the input may come from different sources:

const text = "e\u0301".normalize("NFC");
Buffer.from(text, "utf8").toString("base64");   // "w6k=" (Node.js)
new TextEncoder().encode(text);                  // Uint8Array [195, 169]

In browsers the bytes from TextEncoder can go to bytes.toBase64() where it is supported (Uint8Array.toBase64), or through the btoa helper on the JavaScript page.

import base64, unicodedata

s = unicodedata.normalize("NFC", "nai\u0308ve")
base64.b64encode(s.encode("utf-8")).decode()   # 'bmHDr3Zl'

See Base64 in Python for files and streaming.

FAQ