Dictionary Compression Simulator

Dictionary compression is a technique where frequently occurring substrings are replaced by short back-references to a pre-built compression dictionary. Instead of encoding each character individually, the compressor says "use phrase #42 from the dictionary" — a 2-byte reference that replaces a 20-byte string. Algorithms like LZ77, Brotli, and Zstandard all use dictionary coding internally. The key insight is that a shared dictionary — pre-agreed between compressor and decompressor — eliminates the cold-start problem that makes GZIP ineffective for small payloads. Our dictionary compression simulator models this process using n-gram frequency analysis on your sample data.

1. 1Paste sample data and a target string: The sample data is used to build the dictionary — use representative data of the same type as what you want to compress (e.g. past API responses, log lines, HTML templates). The target is the string you want to compress. Both are processed entirely in your browser — no data is sent to any server.
2. 2Click "Simulate Dictionary Compression": The dictionary compression simulator extracts all n-grams (substrings of configurable length) from the sample data, counts their frequency, and ranks them by total bytes saved. A greedy deduplication pass removes redundant sub-phrases. The top N phrases form the dictionary.
3. 3Compare results: The simulator shows four sizes — original, GZIP only, dictionary only, and dictionary + GZIP — with a visual bar chart, stats grid, and a table of the top dictionary phrases with their frequency and bytes saved. GZIP sizes are exact (native CompressionStream API).

• Dictionary Phrases: The top N substrings from the sample data ranked by total bytes saved — (phrase length − 2) × frequency. The −2 accounts for the 2-byte back-reference cost that replaces each phrase occurrence.
• Replacements Made: The total number of phrase occurrences replaced in the target string. More replacements means the target data closely matches the sample data — a good sign for dictionary effectiveness.
• Dictionary Overhead: The total size of the dictionary itself (phrase bytes + 2-byte index per entry). This is a one-time cost shared across all compressed messages — negligible when compressing many payloads.
• GZIP Sizes (Exact):Actual GZIP compressed sizes computed using the browser's native CompressionStream API — both for the original target and the dictionary-pre-processed target.

Standard GZIP builds its own LZ77 back-reference table from scratch for each compressed payload. For large files this works well — the compressor has enough data to find repeated patterns. But for small payloads(under 1 KB), GZIP's cold-start overhead dominates and compression ratios are poor. A shared dictionary solves this by pre-loading the compressor with known patterns before it sees the target data. This is why Brotli achieves better compression than GZIP for web content — its built-in static dictionary of 13,000+ HTML, CSS, and JavaScript strings gives it a head start on every response. The dictionary compression simulator lets you build and test your own domain-specific dictionary for the same effect.