This tool checks whether two hashes are identical without you having to read them character by character. Use it to compare the checksum a vendor publishes with the one you computed, or two certificate fingerprints shown in different formats.
Before comparing, both values are normalised: converted to lowercase, stripped of spaces, line breaks and colons, and cleared of a leading algorithm prefix such as "sha256-" or "sha256=". As a result, "AB:CD:EF" and "abcdef" are treated as equal, and so are "sha256=2cf24d…" (GitHub's format) and "2cf24d…". The length of the normalised value is then used to estimate the algorithm: 32 hex characters for MD5, 40 for SHA-1, 64 for SHA-256, 96 for SHA-384 and 128 for SHA-512. This detection is only a probability, since SHA3-256 or BLAKE2s also produce 64 characters.
The comparison runs in constant time: the tool goes through every character instead of stopping at the first mismatch, a standard precaution so that timing does not reveal where two secrets differ. When the values do not match, the position of the first differing character is shown, which helps spot a truncated copy or a selection mistake. Different lengths often indicate two different algorithms, for example a SHA-1 compared with a SHA-256. A match only proves that your file is identical to the one used to produce the reference: trust depends on where that reference came from, ideally an official HTTPS page or a signed file.