Decode Base64 to PDF: Recovering a Document From a Payload

Strip the prefix first

If the string starts with data:application/pdf;base64,, everything up to and including the comma is a data URI wrapper, not part of the payload. Decoding with it included produces a corrupt file.

A correct PDF payload begins with JVBERi0, which is Base64 for %PDF-. That prefix is the quickest way to confirm you have a PDF before decoding anything.

Decoding

# command line
base64 -d payload.b64 > document.pdf

# from a JSON field
jq -r '.pdf' response.json | base64 -d > document.pdf

// browser — Blob, not a data URI
const bytes = Uint8Array.from(atob(b64), (c) => c.charCodeAt(0));
const url = URL.createObjectURL(new Blob([bytes], { type: 'application/pdf' }));
window.open(url);
URL.revokeObjectURL(url);

Verifying the result

A valid PDF starts with the bytes %PDF- and ends with %%EOF. Check both — a file that opens the first page and then errors is usually truncated at the end:

head -c 8 document.pdf     # expect %PDF-1.x
tail -c 8 document.pdf     # expect %%EOF

Truncation is the most common failure, and it usually happens upstream — a JSON string field cut to a column width, or a log line clipped. Compare the payload length against the expected file size times 1.33.