How does git hash an object?

Git prepends a header of the form type SPACE length NUL to the content, then takes the SHA-1 of the whole thing. For a blob the header is the word blob, a space, the byte length, and a null byte, followed by the file content.

Why is my hash different from a plain SHA-1 of the file?

Because git hashes the header plus the content, not the content alone. The length is the byte length of the content, and a null byte separates the header from the body, so the input to SHA-1 is longer than the raw file.

Does whitespace and trailing newline matter?

Yes, completely. Git hashes the exact bytes, so a trailing newline, a missing one, or a different line ending all change the length and therefore the hash. Paste content exactly as it would be stored.

What does this match on the command line?

For a blob it matches git hash-object on a file with the same bytes. The commit and tag types let you reproduce the SHA-1 of a commit or annotated tag object given its canonical body text.

Is my content uploaded?

No. The SHA-1 is computed entirely in your browser with a local implementation, so nothing you paste is sent to a server.

Git Object Hash Calculator

The git object hash calculator reproduces the exact SHA-1 identifier that git assigns to an object. Git does not hash file content directly: it first wraps the content in a tiny header, then hashes header-plus-content. This tool builds that header and runs SHA-1 so you get the same 40-character hash as git hash-object.

How it works

Every git object is stored as the bytes type SP length NUL content, where type is blob, tree, commit or tag, SP is a space, length is the decimal byte length of the content, and NUL is a single zero byte (\0). Git computes the SHA-1 of that whole sequence and uses the hex digest as the object’s name.

For example, an empty blob has content of length 0, so the hashed bytes are blob 0\0 — which always produces e69de29bb2d1d6434b8b29ae775ad8c2e48c5391, a hash any git user has seen. The calculator encodes your text as UTF-8 to get the byte length, prepends the header, and runs a self-contained SHA-1 implementation so the result is deterministic and offline.

Example

Content: "hello\n"  (6 bytes)
Header:  blob 6\0
SHA-1 of header+content = ce013625030ba8dba906f756967f9e9ca394464a

That is exactly what printf 'hello\n' | git hash-object --stdin returns.

Notes

Byte length, not character count, drives the header — multi-byte UTF-8 characters and trailing newlines change it, so paste the content precisely. The blob type covers file hashing; commit and tag reproduce those object hashes from their canonical body text. Everything runs locally; your content is never uploaded.