Word List Snapshot Comparison

A player's guide to the language snapshots featured on our website, detailing word rules, tile totals, and how different configurations handle accents in website lookups.

This page compares eight language snapshots generated for the WordSalvo website. The selected website language changes which website list is checked and how the website lookup treats accents. Because languages have different structures, a larger word list does not mean a better dictionary. Please note: these website snapshots are static reference files, so a checker result here is not a promise that a word is in the current live mobile game.

Which Languages Are Included?

This comparison covers the eight language snapshots currently available for reference on our website: English, Dutch, German, French, Spanish, Italian, Norwegian Bokmål, and Turkish. While the WordSalvo mobile game features ten supported languages, including Portuguese and Polish, those two are not yet available in our website's public data.

It is natural to wonder which language has the biggest dictionary, but a larger word count does not result in a better game experience. Each language has unique grammar rules, configured length limits, and source texts. Because of these differences, comparing dictionary sizes is not a fair measure of quality, and we do not declare any language the size winner.

How to Read the Statistics

Each snapshot is labeled with a version string based on its creation date and raw size. 'Source rows' is the original number of lines in the text file, while 'Excluded' shows lines containing numbers or symbols that we remove. 'Normalization duplicates' are words that look different originally (like uppercase versus lowercase) but become the exact same word once our configuration rules are applied, leaving us with the final 'Normalized words' count.

'Max length' reveals the longest word in that specific snapshot. 'Tiles' records the configured number of letter tiles for a standard game bag, and 'Bingo' represents a recorded 45-point configuration value for a seven-tile play. The 'Diacritics' column explains whether accents matter during a website lookup, and the 'words.txt SHA-256' is a unique digital fingerprint we use to verify the file has not been accidentally altered.

Understanding Normalization and Accents

Normalization is the process of making raw dictionary text ready for website lookup. A big part of this is how we handle diacritics, like accents or umlauts. For example, in French, Dutch, Italian, and English, accents are ignored in the lookup, meaning they map to standard, unaccented characters. In contrast, Spanish, German, Norwegian Bokmål, and Turkish require specific accented characters and are kept distinct in the lookup exactly as they appear in the language.

When configuration rules strip away capitalization or ignore accents, multiple distinct words from the original source might end up spelled the exact same way in the list. We call these 'normalization duplicates.' For instance, the Italian and Dutch data show thousands of these duplicates where words that originally had different accents or capital letters collapse into one single list entry.

Website vs. Mobile App Differences

The dictionary data shown on this page was copied from our game on August 15, 2026. Because these are static snapshots, they may not perfectly match the live, up-to-date dictionaries currently active in the WordSalvo mobile app.

We also have a known difference between the English word list on the mobile app and the English statistics displayed on this website. The English totals here represent only the website's specific snapshot.

Finally, this public comparison is for reference only. It does not evaluate which language is best, attempt to cover competitor word games, or guarantee that every term listed here is currently in the mobile game.

Frequently Asked Questions

Why are there only eight languages listed?
Our website currently provides data for eight languages. While the WordSalvo mobile game features ten supported languages, including Portuguese and Polish, those two do not yet have data snapshots available on this website.
Which language has the largest dictionary?
Dictionary size depends on a language's unique grammar, how words are combined, and configured length limits. A larger word list does not mean a better game experience, so we do not declare a size winner or use size as a measure of dictionary quality.
Are these the exact dictionary counts used in the live app?
No. These are static website snapshots. The English dataset has a known difference from the live app, and because these files were copied at a specific time, they may not perfectly match the words currently in the live mobile game.
How are accents and diacritics handled in Spanish and German?
In Spanish, German, Norwegian Bokmål, and Turkish, accents and special characters are kept distinct in the lookup. This is different from English or French, where accents are ignored in the lookup and map to standard unaccented characters.
Where can I find the raw JSON data for these statistics?
Developers can access the aggregate snapshot statistics through the /research/dictionaries.json endpoint on this website.
Technical Appendix: Data Matrix

The following table provides detailed technical metrics for the specific snapshot versions hosted on the website, explaining how raw source files are converted into website snapshot data.

LanguageVersionNormalized wordsSource rowsExcludedNormalization duplicatesMax lengthTilesBingoDiacriticswords.txt SHA-256
English2026-06-14-284000284,000284,000001510445Ignored for lookup5adb53d1687c27a10b29a43440933aee97b887cde6c19bd962a26375d1c134a2
Nederlands2026-06-14-465552462,773465,55262,7732110445Ignored for lookup29b578e9361fbfefdf787d6787d20234445103af315e1021f582df13d45be218
Deutsch2026-05-27-588414588,414588,414001510245Kept separate0e041e1f32706abf887f55323a6c42d4bbd8c1161d3addc46296d0b514aea403
Français2026-06-14-403464403,464403,464001510245Ignored for lookup50452f62e460d31eebb38a51dbe099e7d0116e8f3c09df9a6765e7be3d4155dc
Español2026-05-30-588265588,265588,265001610045Kept separate7915888b764f2c1b87bba638a22ed1edd894b6dda6f408a38b11cc8e9819b099
Italiano2026-06-14-550050530,542550,050619,5021812045Ignored for lookupfdc5e6793bb58f61273675df2959538709f3df87fe0ed1caf7c9e11bd8f7eda5
Norsk2026-05-27-812569812,569812,569001510045Kept separate37ea28a8496aee3327e8ec9a6facd2825b56b19617797c07c09725f519886554
Türkçe2026-06-14-155258155,258155,258001510045Kept separatecb40ed08b2a9f5f2b1514e7d108d95d18e3e9a3ffa7209ef4dfbf7e3fd089503

Technical Appendix: Reproducing the Data

Developers and researchers can verify these snapshot metrics locally using our website repository's statistics script.

  1. 1Ensure you have a local clone of the website repository and navigate to the project root directory.
  2. 2Run the reproduction command: npm run word-tools:stats.
  3. 3The script reads the public/word-data/<locale>/glosses.json metadata and the words.txt file for each language, then regenerates the statistics and sparse lookup JSON.

Additional Resources

Want to see if a specific term appears in the website list? Visit /en/tools/word-checker to test specific words, or go to /en/corrections to suggest missing vocabulary. You can also read about our full process at /en/research/dictionary-methodology or download the raw technical data directly at /research/dictionaries.json.