About this charset converter
Type any text and four representations update in real time: UTF-8 byte hex (two digits per byte, space-separated), percent-encoding (same as encodeURIComponent), HTML numeric entities (one &#xNNNN; per code point) and Unicode escapes (\uXXXX, with characters outside the BMP automatically split into surrogate pairs). Chinese, emoji and full-width characters are all handled per code point — handy for debugging API payloads, tracking down mojibake, or writing regex and escaping configs.
Decoding works the other way too: paste any of the formats above into the decode box, hit Decode, and the format is auto-detected (hex byte string / %XX / &#x…; / \uXXXX) and converted back to text, with the detected format shown next to the button. Note that plain words made of hex letters (like deadbeef) are naturally shaped like hex byte strings and will be decoded byte-wise — an inherent ambiguity of auto-detection. Everything runs locally; nothing is uploaded.