Question 1

How do I convert Base64 to a PDF?

Accepted Answer

Paste the Base64 (with or without the data:application/pdf;base64, prefix) and check that the detected type is PDF. A PDF in Base64 starts with JVBERi, which is %PDF- encoded. The tool shows the PDF in the preview and downloads it with a .pdf extension. If the API response is JSON, paste the whole JSON: it finds the longest field that looks like Base64 and lets you switch fields.

Question 2

How can I tell what type of file an unknown Base64 is?

Accepted Answer

Look at the start of the text. JVBERi is PDF, iVBORw0KGgo is PNG, /9j/ is JPEG, R0lGOD is GIF, UEsDB is ZIP (or DOCX, XLSX, PPTX), UklGR is a RIFF container (WebP or WAV), H4sI is GZIP and SUQz is MP3 with an ID3 tag. The tool reads this from the decoded bytes and, for ZIP, looks at entry names to separate the Office formats.

Question 3

What is the difference between Base64, base64url and a data URI?

Accepted Answer

Standard Base64 uses + and / and ends with = when the last group of 4 characters needs completing. base64url (RFC 4648, section 5) swaps + for - and / for _ and usually drops the =, which is why it fits URLs and JWTs. A data URI (RFC 2397) is Base64 with a data:type;base64, prefix to embed a file in HTML or CSS. The tool reads all three and picks the alphabet on its own.

Question 4

What does the invalid character error mean?

Accepted Answer

The text has something outside the alphabet, such as quotes, a comma, an accent, a literally written 
 or an invisible space. The tool reports the position, line, column and Unicode code of the character. Line breaks, spaces, URL-encoded %2B and %3D and JSON escapes are fixed automatically; whatever is left becomes an error, and you can decode ignoring the invalid ones.

Question 5

Does Base64 work without the equals sign at the end?

Accepted Answer

It does. Padding only completes the last group of 4 characters: 2 final characters are worth 1 byte and 3 are worth 2 bytes. A remainder of 1 character is impossible, because 6 bits do not make a byte, and the tool reports a length error. An = in the middle of the text means two Base64 strings were pasted together and is flagged too.

Question 6

How much does Base64 increase the file size?

Accepted Answer

About 33%: 3 bytes become 4 characters, or 4 x ceil(n/3). A 1,000,000 byte file becomes 1,333,336 characters. With MIME line wrapping (76 columns and CRLF) it adds 2 bytes per line and reaches 1,368,422 bytes. Without padding the text loses up to 2 characters. That is why embedding large files in JSON or a data URI weighs more than sending them as binary.

Question 7

Is the preview safe with unknown files?

Accepted Answer

Yes. The type comes from the decoded bytes, not from what the input declares. Images and SVG appear only through an image tag, which does not run scripts. HTML, XML and JSON appear as text. PDF, audio and video use the browser players, and everything else has no preview, only a download and the first bytes in hexadecimal.

Question 8

How large can the input be?

Accepted Answer

Decoding and encoding work in chunks and hand control back to the browser, so the page does not freeze. The practical limit is tab memory: the tool accepts text files with Base64 of up to 200 MiB and encodes files of up to 300 MiB. Output goes to the clipboard only up to 20 million characters; above that use the download.

Question 9

Does it unzip GZIP or open the ZIP?

Accepted Answer

It recognizes GZIP by the bytes 1F 8B and downloads the file as .gz, but does not decompress it. For ZIP, DOCX, XLSX and other ZIP-based files it lists the central directory entries (name, size and method) without extracting the content. To extract, download the file and open it on your system.

Question 10

Is my file sent to any server?

Accepted Answer

No. Decoding, detection, the SHA-256 and the preview all happen in the browser, using local Blobs. Even so, avoid pasting private keys or sensitive documents into any site that is not your own trusted environment.

Base64 to File Decoder (PDF, Image, ZIP)