How it works
Numeric references, both decimal (é) and hexadecimal (é), are converted directly. Named references are looked up in a table of 175 of the most common entities, covering punctuation, symbols, arrows, Greek letters and Western European accented letters. The input is only read as text; nothing is rendered.
Rules this tool follows
- Named entities must end with a semicolon; “&” without one is left as typed.
- Invalid code points, such as �, become the replacement character �.
- References in the range €–Ÿ are mapped the way browsers do (Windows-1252), so – becomes an en dash.
- Decoding runs once; double-encoded text (&) needs a second pass.
Example
Input
Tom & Jerry’s café © 2026
Output
Tom & Jerry’s café © 2026
Common uses
- Cleaning text copied from HTML source, RSS feeds or APIs.
- Reading escaped content in databases and logs.
- Converting entity-heavy exports into plain text.
Limitations and notes
- Rare named entities outside the built-in table are left unchanged and flagged. Numeric references always work.
- To remove HTML tags as well, use HTML to Text.