Identify the true type of a file by its magic-number byte signature — not the (easily-spoofed) filename extension.
Upload any file and the tool reads its leading bytes, then matches them against a database of ~70 magic-number signatures covering images, documents, audio, video, fonts, archives, executables and more. It reports the detected type, MIME type, recommended extension, and a confidence score reflecting how specific the match is.
How it differs from magic-number-detector — that tool takes a hex string you type by hand and looks it up. This tool reads the actual uploaded file directly, supports offset-based signatures (e.g. ZIP-relative markers at byte 4), applies bitmasks (e.g. the RIFF-size field is ignored), and surfaces confidence + all candidates when multiple signatures match.
What it catches
- A .png that's actually a JPEG (mismatch warning).
- An executable renamed .txt.
- A WebP hidden inside a .jpg.
- Office .docx / .xlsx (which are really ZIP containers).
- Mach-O / ELF / PE executables, Java .class, SQLite databases, fonts, etc.
Confidence
- ≥90%: long, anchored signature (e.g. PNG's 8-byte sequence at offset 0) — near-certain.
- 70-89%: solid match but shorter signature.
- 45-69%: ambiguous (e.g. a single leading byte like '{' for JSON).
- <45%: weak guess.
The hex dump of the first 64 bytes is always shown so you can verify the signature by hand.