Calculate Shannon entropy of text, files, or byte sequences. Visualize entropy distribution to detect encrypted or compressed data.
Paste text, drop in a file, or type a byte sequence, and this tool reports the input's Shannon entropy in bits per byte: how evenly its byte values spread across the 0-255 range, from 0 (one byte repeated) to 8 (all values equally likely). A distribution view plots each byte value's share of the input, so encrypted or compressed payloads stand out as a flat band near the top. The thing most users get wrong: entropy ignores order, so a repeating pattern and a shuffled string made of the same characters score identically. A high number means no byte-level regularity, nothing deeper.
A concrete input and expected output from the current implementation.
Input
the quick brown fox jumps over the lazy dog
Expected output
43 bytes, 27 distinct byte values; Shannon entropy 4.39 bits per byte of a maximum 8; total entropy 188.6 bits.
The 43 bytes use 27 distinct values: the space appears 8 times, 'o' 4 times, 'e' 3 times, 't', 'h', 'u', and 'r' twice each, and the remaining 20 letters once each. With N = 43 and count c per symbol, H = -sum(p*log2(p)) = (8*log2(43/8) + 4*log2(43/4) + 3*log2(43/3) + 8*log2(43/2) + 20*log2(43)) / 43 = 188.57 / 43 = 4.39 bits per byte, and total entropy is 43 * 4.39 = 188.6 bits. The eight spaces and repeated letters concentrate probability, pulling the score below the 8.0 a uniform byte stream would show, while most letters appear once, keeping it above 0. Ordinary English prose generally lands between about 4 and 4.5 bits per byte.
The tool tabulates how often each of the 256 possible byte values occurs, converts the counts to probabilities, and computes H = -sum(p*log2(p)), reported in bits per byte. All 256 values equally likely gives exactly 8; a single repeated byte gives 0. The formula is logarithmic, so a mid-range value like 4 is not 'half random' - it reflects a genuine spread among a few dozen symbols.
The distribution view draws one bar per byte value, proportional to how often that value appears, so structure shows at a glance: plaintext clusters in a narrow band of letters and punctuation, while encrypted, compressed, or hashed data spreads evenly across all 256 positions. Files are read byte-for-byte as stored, so container headers, checksums, and framing count as ordinary bytes and can pull an otherwise high-entropy payload's score down.
There is no clean threshold. Encrypted and well-compressed data typically score from roughly 7.9 up to just under 8 bits per byte on a large enough sample, with every byte value appearing about equally often, while plaintext and structured data score lower. But base64, images, hashes, and packed formats also reach the high end, so the number alone cannot prove encryption - it only shows the data has no byte-level regularity.
Because the tool measures the spread of byte frequencies, not meaning or predictability. Letters and the space character appear with naturally uneven frequencies, and that spread works out to roughly 4 to 4.5 bits per byte for ordinary prose. The often-quoted figure of about 1 bit per character comes from predictive letter models that use surrounding context, which a pure frequency count never does.
The example and behavioral notes were checked against the browser implementation. Standards and primary references below define the relevant format, formula, or platform behavior.
Generate strong, random passwords with customizable length, character sets, and complexity requirements.
Analyze password strength with entropy scoring, crack-time estimates, and improvement suggestions.
Create cryptographically random strings for API keys, tokens, salts, and unique identifiers.
Create QR codes for URLs, text, Wi-Fi credentials, vCards, and more with customizable size and colors.
Scan and decode QR codes from uploaded images or your device camera to extract embedded data.
Compute MD5, SHA-1, SHA-256, and SHA-512 checksums to verify file integrity and detect tampering.