Base64 Padding Tool

Input

character . Fill: 0 =

Input

character . Fill: 0 =

Result

{count character =change} = 0

Calculate the correct trailing = padding for any Base64 string per RFC 4648. Add or remove padding in real time to fix strings that do not match the standard. Everything runs locally.

Base64 Encode

Base64 Decode

Base64 Cleaner

Base64 Formatter

URL Safe Base64

Base64 Length Statistics

What is Base64 padding?

Base64 padding is the '=' character defined by RFC 4648. It exists to make every Base64 output length a multiple of 4. Base64 maps 3 bytes (24 bits) into 4 characters, so when the input is not a multiple of 3 bytes the trailing group needs to be padded with = so decoders can tell exactly how many real bytes are present.

Padding rules: ① input bytes % 3 = 0 → 0 trailing =; ② % 3 = 1 → 2 trailing =; ③ % 3 = 2 → 1 trailing =. Equivalently, the Base64 character count % 4 directly tells you how many = are missing.

Why padding matters: Base64 decoding needs an explicit end. Without padding, a 1-byte input and a 2-byte input can produce Base64 outputs of ambiguous length, making it impossible for a decoder to tell how many real bytes were encoded.

In practice: most protocols (Data URL, MIME, JWT, config files) strictly require padding. A few scenarios (URL paths, file names, short links) omit padding to save bytes. This tool supports both directions.

Use Cases

Fix Base64 decode errors such as InvalidCharacterError or 'length not a multiple of 4' before re-decoding.
Clean up Base64 copied from APIs, JWT tokens, or logs that contain extra whitespace or missing padding.
Generate Data URLs whose Base64 portion strictly conforms to RFC 4648 so browsers and third-party tools decode it consistently.
Quickly validate and pad Base64 for JWT tokens, config files, and API credentials.
Strip padding to save bytes when the target protocol supports unpadded Base64 (URL paths, file names).
Learn Base64 encoding by comparing input byte count with required padding count.

How to Use

Paste or type the Base64 string you want to process (with or without padding, with or without newlines).
Choose a mode: add (append = to the end) or remove (strip trailing =).
The tool calculates and shows the result instantly, with character count and padding delta displayed live.
Copy the result to the clipboard or download it as a .txt file for downstream use.

Features

Auto-calculates padding count: shows the required = count versus the current one in real time.
Add mode: pads any Base64 string to a multiple-of-4 length to fix length errors in one click.
Remove mode: strips trailing = characters for URL paths, file names, and other space-sensitive scenarios.
Ignores newlines and whitespace: paste Base64 copied from emails, JWT output, or config files directly.
Live character count: shows input length, output length, and the exact number of = added or removed.
One-click copy and download: copy the result to the clipboard or save it as a .txt file.
Local browser processing: every calculation runs in your browser; the original input never leaves your device.

Best Practices

Verify the input is real Base64 before padding

This tool only appends trailing =; it does not clean up invalid characters. If the input contains spaces, CJK characters, or symbols outside the Base64 alphabet, decoding will still fail after padding. Use a Base64 cleanup tool first to strip non-alphanumeric characters.

Always pad for Data URL and JWT

Data URL (RFC 2397) and JWT (RFC 7519) strictly require padding. Omitting = triggers decode errors or inconsistent results in most implementations. Always use this tool to pad to a multiple of 4 before submitting.

URL-safe Base64 requires more than stripping padding

If you want Base64 inside a URL path, file name, or short link, simply removing padding is not enough. You also need to replace + with - and / with _ (Base64URL). Without that substitution, + and / remain illegal in URLs.

Decoder still complains after removing padding

The target tool probably requires strict padding. Switch to add mode and try decoding again. If it still fails, the input itself may have been corrupted before stripping.

Very large Base64 strings: prefer the command line

This tool is best for single Base64 segments. For tens of megabytes of Base64, prefer command-line tools such as openssl base64 or the base64 utility to avoid browser slowdowns.

Teaching Base64 padding always shows the table

Padding is the most confusing part for newcomers. When explaining, attach the 'input bytes mod 3 → = count' table (see referenceTables on this page). Without it, students cannot tell why some inputs need 1 = while others need 2.

FAQ

What is Base64 padding for?

It makes every Base64 output length a multiple of 4 so the decoder has an explicit end. Without padding, 1-byte and 2-byte inputs can produce same-length Base64 strings, leaving the decoder unable to tell the real byte count.

Why must Base64 length be a multiple of 4?

Because Base64 maps every 3 bytes into 4 characters. When the input is not a multiple of 3 bytes, the trailing group is padded with = to reach 4 characters. Any Base64 string whose length is not a multiple of 4 is malformed.

How do I calculate the padding count?

Use (4 - input_bytes % 3) % 3, or look at the Base64 character count mod 4: remainder 0 means 0 =, 2 means 1 =, 3 means 2 =. This tool does the calculation automatically.

Can I just remove the padding?

Yes, but only when the target protocol supports unpadded Base64. File names, URL paths, and some JWT implementations accept it; Data URLs, MIME, and most config files require padding. Use this tool's remove mode only when you have confirmed the target accepts it.

Decoder says 'length is not a multiple of 4', what do I do?

Switch to add mode in this tool to append the missing =. If the error persists, the input itself contains invalid characters and needs to be cleaned with a Base64 cleanup tool.

Does adding padding change the original bytes?

No. Padding only appends = characters at the end. The original bytes in the middle are untouched, so decoding the padded string returns exactly the same byte sequence.

Does it handle Base64 with newlines or spaces?

Yes. The tool strips newlines, carriage returns, and spaces before computing padding. Multi-line Base64 from email attachments, JWT output, or config files can be pasted directly.

Is padding the same as Base64URL?

No. Padding is the trailing = issue; Base64URL is a character-set issue (replacing + with - and / with _). This tool only handles padding. Use a dedicated Base64URL tool for URL-safe conversion.

Is my input uploaded to a server?

No. Every padding calculation runs locally inside the browser. The original Base64 string never leaves your device, so you can safely process credentials and tokens.

Troubleshooting

Decoder says 'length is not a multiple of 4'

Use this tool's add mode to append the missing = and try decoding again. If the error persists, the input itself contains invalid characters; run it through a Base64 cleanup tool first.

Input might contain invalid characters

Standard Base64 contains only A-Z, a-z, 0-9, +, /, =. Any other character (CJK, symbols outside the alphabet) is illegal. This tool strips whitespace automatically, but other invalid characters need a separate Base64 cleanup step.

URL still fails after removing padding

The URL may still contain characters that need URL encoding (?, &, =), or it may still contain + and / from the Base64 alphabet. Stripping padding alone is not URL-safe; you also need Base64URL conversion or URL encoding.

No result after pasting

The input may consist entirely of whitespace (newlines, spaces, tabs). The tool ignores whitespace when computing padding, but an entirely blank input produces no result. Paste at least one Base64 character.

Glossary

Padding = character: Trailing character defined by RFC 4648 used to indicate how many real bytes are missing from the last Base64 group. Valid padding only appears at the end of the string.
Multiple of 4: RFC 4648 requires Base64 output length to be a multiple of 4, otherwise it is considered invalid encoding. Strict decoders raise an error immediately.
Padding formula: Required = count = (4 - input_bytes % 3) % 3. Equivalently, base64 length mod 4: remainder 0 means 0 =, 2 means 1 =, 3 means 2 =.
Base64 without padding: A Base64 variant that omits the trailing =. Common in URL paths and file names, but technically not RFC 4648 compliant and may fail in strict decoders.
Base64 character set: A-Z, a-z, 0-9, +, / (64 characters) plus = for padding. Any other character is invalid and must be cleaned up with a Base64 cleanup tool first.

Base64 padding rules at a glance

Mapping between input byte count and required trailing = padding.

Input bytes	mod 3	Base64 length	= padding	Example (input → output)
`3n`	0	`4n`	0	ABC → QUJD
`3n+1`	1	`4n+2`	2	AB → QUI=
`3n+2`	2	`4n+3`	1	A → QQ==

Common protocols and padding requirements

Different protocols treat = padding differently. Removing it where it is required causes decode failures.

Scenario	Padding required	Typical use
`Data URL (RFC 2397)`	Recommended (compatibility)	HTML / CSS / inline images
`JWT (RFC 7519)`	Required (strict)	OAuth / API tokens
`MIME (RFC 2045)`	Required	Email attachment encoding
`URL paths / file names`	Optional (commonly stripped)	Short links / cache keys
`Config files (YAML / JSON)`	Required	Credentials, signed values