Base64 Size Calculator

How to calculate Base64 size

1. Keep Bytes → Base64 length selected and type the Original size, choosing the Unit. Under Units, pick 1 KB = 1000 B or 1 KiB = 1024 B to match how your size was measured.
2. Or drop a file on the box. Only its size is read; the file’s content is never opened.
3. Read the table: padded and unpadded length, the overhead in percent, a custom wrap (set Line length and Line break), MIME email lines, a PEM body, a data: URI with the file’s MIME type and a JSON string with quotes.
4. For the other direction, switch to Base64 length → bytes, enter the length, say whether it includes line breaks, and choose the Padding case: known, unknown (0–2 ×) or none. Pasting the Base64 itself gives an exact answer.

The live formula under the table repeats the calculation with your own numbers, which is handy for a code comment or a ticket.

The formula and why it is 33%

Base64 turns every 3 bytes into 4 characters, each carrying 6 bits. The last group is padded with = to a full 4 characters, so:

padded   = 4 × ⌈n ÷ 3⌉
unpadded = ⌈4n ÷ 3⌉
decoded  = (characters × 3 ÷ 4) − number of "=" signs

For large inputs that is a fixed 33.3% increase. For tiny ones the padding dominates: one byte becomes four characters, an overhead of 300%. The table below was produced with the same code the calculator runs:

Line breaks are counted between lines, with none after the last line. Some encoders, Python’s base64.encodebytes for example, end every line including the last with \n, which adds one more break. Add the prefix for a data URI: data:image/png;base64, is 22 characters.

Planning around size limits

Typical limits worth checking before something fails:

• Request body limits. If an endpoint accepts 10,000,000 bytes and the file goes in a JSON field, the file can be at most about 7,500,000 bytes, minus the rest of the JSON. Switch to reverse mode, enter the limit, and subtract your field names.
• Database columns and environment variables. A VARCHAR(255) holds Base64 of at most 189 bytes; 190 bytes already need 256 characters.
• Email. MIME attachments are wrapped at 76 characters with CRLF, so they grow by about 37% in total. A message size limit therefore admits files only about 73% of its size. Base64 in email has the details.
• Data URIs in HTML or CSS. Every inlined image costs a third more than the file and cannot be cached separately. Data URI performance explains when it still pays off.

Base64 length in code

Every language can compute it without encoding anything:

import math

def b64_len(n, padded=True):
    return 4 * math.ceil(n / 3) if padded else math.ceil(4 * n / 3)

print(b64_len(1048576), b64_len(1048576, padded=False))

That prints 1398104 1398102. In JavaScript the padded length is 4 * Math.ceil(n / 3). To check a string you already have, len(s) * 3 // 4 - s.count("=") gives the decoded size, after you strip line breaks. If reverse mode reports an impossible length, such as a remainder of 1 after dividing by 4, the string is damaged: run it through the Base64 validator to find out where.

FAQ