User Rating 0.0 ★★★★★

Total Usage 0 times

Category File Converters

Drop a .txt file here or click to paste from clipboard

Base64 String or Data URI 0 characters Supports raw Base64, Data URIs, and URL-safe Base64

File Name (optional)

Is this tool helpful?

Your feedback helps us improve.

★ ★ ★ ★ ★

About

Base64 encoding inflates binary data by approximately 33%, converting every 3 bytes into 4 ASCII characters drawn from an alphabet of 64 symbols (A - Z, a - z, 0 - 9, +, /). This tool reverses that process. Paste a raw Base64 string or a full Data URI and the decoder reconstructs the original binary, identifies the file type via magic-byte signatures, and delivers a downloadable file. Malformed input or truncated strings produce corrupted output that most applications will silently reject. Verify your source encoding before transmission.

MIME detection works in two passes: first the data: URI prefix is parsed if present; then the first 16 decoded bytes are matched against known file signatures. The tool handles images, documents, audio, video, fonts, and archives. For text-based formats (JSON, XML, CSV, plain text), a UTF-8 heuristic is applied. Note: Base64 strings without a Data URI header default to application/octet-stream when no magic bytes match. This is correct behavior per RFC 2045 §6.8.

Formulas

Base64 encoding maps each group of 3 input bytes (24 bits) to 4 output characters from a 64-character alphabet. The encoded size relates to the original by:

S_encoded = 4 ⋅ S_original3

Where the division rounds up to the next integer (ceiling). The inverse operation recovers the original byte count:

S_original = 34 ⋅ S_encoded − P

Where S_encoded = length of the Base64 string (excluding whitespace), S_original = decoded byte count, and P = number of padding = characters (0, 1, or 2). The size overhead ratio is constant at 43 ≈ 1.333, meaning a 33.3% increase. Each Base64 character encodes exactly 6 bits: log₂(64) = 6. The character set per RFC 4648 is: A - Z (indices 0 - 25), a - z (26 - 51), 0 - 9 (52 - 61), + (62), / (63). URL-safe variant replaces + with - and / with _.

Reference Data

File Format	Magic Bytes (Hex)	MIME Type	Typical Extension	Base64 Prefix Chars
PNG	89 50 4E 47	image/png	.png	iVBOR
JPEG	FF D8 FF	image/jpeg	.jpg	/9j/
GIF	47 49 46 38	image/gif	.gif	R0lG
WebP	52 49 46 46 .. .. .. .. 57 45 42 50	image/webp	.webp	UklG
BMP	42 4D	image/bmp	.bmp	Qk
TIFF (LE)	49 49 2A 00	image/tiff	.tiff	SUkq
TIFF (BE)	4D 4D 00 2A	image/tiff	.tiff	TU0A
ICO	00 00 01 00	image/x-icon	.ico	AAAB
SVG	3C 73 76 67 or 3C 3F 78 6D	image/svg+xml	.svg	PHN2
PDF	25 50 44 46	application/pdf	.pdf	JVBE
ZIP	50 4B 03 04	application/zip	.zip	UEsD
GZIP	1F 8B	application/gzip	.gz	H4sI
RAR	52 61 72 21	application/x-rar	.rar	UmFy
7Z	37 7A BC AF	application/x-7z	.7z	N3q8
MP3 (ID3)	49 44 33	audio/mpeg	.mp3	SUQz
WAV	52 49 46 46 .. .. .. .. 57 41 56 45	audio/wav	.wav	UklG
OGG	4F 67 67 53	audio/ogg	.ogg	T2dn
FLAC	66 4C 61 43	audio/flac	.flac	Zkxh
MP4	.. .. .. .. 66 74 79 70	video/mp4	.mp4	AAAA (varies)
WebM	1A 45 DF A3	video/webm	.webm	GkXf
WOFF2	77 4F 46 32	font/woff2	.woff2	d09G
WOFF	77 4F 46 46	font/woff	.woff	d09G
WASM	00 61 73 6D	application/wasm	.wasm	AGFz
EXE/DLL	4D 5A	application/x-msdownload	.exe	TVo
SQLite	53 51 4C 69	application/x-sqlite3	.db	U1FM
JSON (heuristic)	7B or 5B	application/json	.json	eyJ / W
XML	3C 3F 78 6D	application/xml	.xml	PD94
Plain Text	UTF-8 heuristic	text/plain	.txt	Varies

Frequently Asked Questions

The browser's native atob() function will throw an InvalidCharacterError for strings with illegal characters or incorrect padding. If the string decodes but was truncated, the resulting file will be shorter than expected. For images, this typically manifests as a partially rendered image with a gray or black band at the bottom. For ZIP archives and PDFs, the file will be unreadable. Always verify the decoded file size matches the expected original: S_original = 34 ⋅ S_encoded − P.

Yes. The converter automatically detects and normalizes URL-safe Base64 variants per RFC 4648 §5. Characters - are replaced with + and _ with / before decoding. Missing padding (=) is also restored. This covers tokens from JWTs, URL parameters, and data transmitted through systems that prohibit + and /.

When no data:[mime];base64, prefix is present, the tool decodes the first 16 bytes and compares them against a signature database of known magic bytes. For example, PNG files always begin with 89 50 4E 47 0D 0A 1A 0A. MP4 files contain 66 74 79 70 at byte offset 4. Text-based formats (JSON, XML) are identified via UTF-8 validity heuristics and opening character analysis. If no signature matches, the file defaults to application/octet-stream.

The practical limit depends on browser memory. Strings exceeding approximately 100MB decoded may cause tab crashes. For strings larger than 1MB (approximately 1.33MB encoded), this tool offloads decoding to a Web Worker to keep the UI responsive. Chrome's V8 engine has a string length limit of approximately 512MB (2²⁹ characters). Firefox's limit is similar.

Three common causes: (1) The Base64 string contains line breaks or whitespace that inflate the character count but are stripped before decoding. (2) The string includes a Data URI prefix which is metadata, not payload. (3) Padding characters (=) were stripped during transmission. The tool handles all three cases by cleaning the input before processing. The true decoded size is always 34 of the cleaned Base64 character count, minus padding bytes.

Yes. Base64 is format-agnostic. Any binary encoded as Base64 decodes back to its original bytes. The tool recognizes signatures for EXE/DLL (4D 5A), WebAssembly (00 61 73 6D), SQLite (53 51 4C 69), and dozens of others. The decoded file downloads with the correct extension. Verification is your responsibility: always check file integrity with a hash comparison when available.