UTF-8 vs UTF-16

Byte sizes

• ASCII (A–Z, digits, punctuation): 1 byte in UTF-8, 2 in UTF-16.
• Latin with accents, Greek, Cyrillic, Hebrew, Arabic: 2 bytes in both.
• Chinese, Japanese, Korean: 3 bytes in UTF-8, 2 in UTF-16.
• Emoji and other astral characters: 4 bytes in both.

Why UTF-8 won the web

Backward compatibility. Any ASCII file is already valid UTF-8, byte for byte, so decades of existing text and protocols kept working with no migration. UTF-8 also has no byte-order ambiguity — UTF-16 comes in big-endian and little-endian forms and needs a byte-order mark to tell them apart — and it never produces a zero byte inside a character, which would terminate a C string early. Around 98% of web pages now use it.

Surrogate pairs, and why lengths disagree

UTF-16 represents characters above U+FFFF as two 16-bit units called a surrogate pair. Languages that expose strings as UTF-16 units — JavaScript, Java, C# — therefore report a length of 2 for a single emoji. In Python 3 or Go, the same character counts as one. Neither is wrong; they are counting different things. It is why "😀".length is 2 in JavaScript, why slicing a string can split a character in half, and why iterating by code point rather than by index matters.

Where each is used

UTF-8 for files, HTTP, JSON, XML, databases, source code — anything stored or transmitted. UTF-16 mostly appears as an in-memory representation: JavaScript strings, Java and C# strings, the Windows API, and some older file formats. You rarely choose UTF-16 deliberately; you encounter it because a platform uses it internally.

Practical advice

Declare UTF-8 everywhere — HTTP headers, HTML meta tags, database and connection charsets, and the encoding argument wherever files are read or written. Almost all garbled text comes from one link in that chain disagreeing with the others. When counting or slicing strings, count what you actually mean: bytes for storage limits, code points for characters, and grapheme clusters for what a person would call a character.

Frequently asked questions

• Unicode Escape and Unescape Convert characters to \uXXXX escapes and back. In your browser
• Character Counter Count characters with and without spaces, live as you type. In your browser
• Text to Binary Converter Convert text to binary, hex, octal or decimal and back. In your browser
• Text Cleaner Strip formatting, smart quotes and invisible characters. In your browser
• Base64 Encoder and Decoder Encode text to Base64 or decode it back, including URL-safe. In your browser

• How to Fix Broken Characters in Text Diagnose mojibake, question marks and invisible characters, and repair the text.
• Base64 vs Encryption Base64 is an encoding, not a cipher. What it is actually for, and why it protects nothing.