Generate text digests with SHA-256, SHA-384, SHA-512, SHA-1 or MD5 locally. Choose hexadecimal or Base64, inspect the UTF-8 byte count and copy or download the digest.
Text is encoded and hashed on your device without uploads, accounts or saved input history. SHA algorithms use the browser’s Web Crypto implementation, while MD5 uses a small self-hosted open-source library. Exports contain only the digest, not the original text.
SHA-256 is the default, with SHA-384 and SHA-512 also available. MD5 and SHA-1 are provided for legacy compatibility. Output can be hexadecimal or standard padded Base64. Hex letter case changes only the representation; Base64 is case-sensitive and is not changed by that option.
Text is UTF-8 encoded with no trimming or Unicode normalization. Spaces, tabs, leading/trailing whitespace and an actual text BOM are retained. CRLF and CR line endings normalize to LF to match browser textarea behavior. A final line break changes the digest. Visually similar composed and decomposed Unicode text can have different hashes. Unpaired surrogate units are rejected instead of silently replaced.
Empty text is supported and produces the algorithm’s standard empty-input digest when you click Hash text. This tool hashes text only; it does not read or hash local files. Clear empties the input and result while keeping options, and does not automatically hash the empty input.
No. A digest is not reversible encryption and this tool does not implement password hashing, salts or key derivation. MD5 and SHA-1 are unsuitable for collision-resistant security applications. A matching digest alone does not authenticate the origin of data. Choose an algorithm and workflow appropriate to your application.
SHA hashing requires Web Crypto in a supported secure context such as HTTPS or localhost. If unavailable, the run fails rather than switching algorithms. MD5 uses self-hosted SparkMD5 3.0.2, with the original license retained and npm package integrity verified. No algorithms are fetched from a CDN or hashing server.
Input is limited to 200,000 UTF-16 text units before line-ending normalization. Errors produce no partial digest. Text and options are not uploaded or stored. Editing while a SHA calculation is pending clears the old result and ignores the late response; the underlying browser operation is not forcibly canceled. Reports contain only the digest. Try an example restores abc with SHA-256 and lowercase hex.
