This is the tool you open when Wireshark hands you a hex dump and you need to know if that payload is a plaintext password, an HTTP request, or just binary noise pretending to be a security incident.
Paste hex, get text. That's it. The reason this page exists is that raw hex is impossible to triage by eye. 48545450 looks like noise until you translate it: HTTP. 69735f6d61 is is_ma. Spot a few tokens like that and you know what you're looking at.
The boring details: two hex digits per byte. The tool forgives you if you separate bytes with spaces, colons, commas, or line-prefix addresses like Wireshark outputs (0x0000: 48 54 54 50). It strips prefixes, ignores whitespace, takes whatever shape your dump arrived in.
The encoding dropdown matters. Default is UTF-8, which handles modern text correctly: emoji, Japanese, accented Latin, Cyrillic, all the multi-byte sequences that make Unicode fun. If you see garbled mojibake in the output bytes above 0x7F, you're probably dealing with Latin-1 or some other single-byte encoding. Switch to it. ASCII works fine for pure English but corrupts anything above code point 127 — a UTF-8 encoded e-acute is two bytes (0xC3 0xA9), which two ASCII conversions read as two garbage characters.
Non-printable bytes show up as escape codes like x00, tab, newline, carriage-return. Important distinction: some editors truncate at null bytes, silently hiding the rest of your dump. The escape notation here doesn't drop them — you see x00 as a visible token and your context is preserved.
For the reverse direction, paste text and get hex. Useful for hand-crafting binary protocol packets or computing expected byte sequences for unit tests. Select the delimiter style the consumer expects.
The "what file is this" trick: grab the first 8-16 bytes of any file, paste them as hex, and look for magic bytes. 89 50 4E 47 is PNG. FF D8 FF is JPEG. 25 50 44 46 is PDF. 50 4B 03 04 is ZIP (and by extension, most modern Office formats). Every file type has a signature and hex is how you read it before any tool will open the file.
Odd-length hex strings will fail — every byte needs two digits. Don't ask me why someone would hand you 47 hex chars instead of 48, but it happens. The tool flags it rather than silently dropping the last nibble. Fixing it is your problem.