Find the secret word through semantic similarity
Welcome to a word game that understands what words mean, not just how they're spelled.
Wordsy is a unique word-guessing game that uses artificial intelligence to understand the meaning of words. Unlike traditional word games that focus on spelling, anagrams, or letter patterns, Wordsy uses semantic similarity—the mathematical representation of how closely related two words are in meaning.
When you guess a word, our AI compares it to a secret target word and tells you how semantically close you are. Words are ranked from 1 (the target) to 50,000+, with closer rankings indicating greater semantic similarity. This creates a game where:
Wordsy is powered by GloVe (Global Vectors for Word Representation), a word embedding model developed by researchers at Stanford NLP. Word embeddings are a technique in natural language processing where words are represented as vectors in a high-dimensional space.
These vectors are learned from massive text corpora containing billions of words. The training process captures semantic and syntactic relationships: words that appear in similar contexts end up close together in vector space. This is why "king" and "queen" are near each other, as are "doctor" and "hospital."
When you submit a guess, our system:
Cosine similarity measures the angle between vectors, giving a value between -1 and 1. Words with higher similarity scores receive better (lower) ranks. This mathematical approach means the game truly understands meaning, not just surface-level patterns.
Our backend uses vectorized NumPy operations to calculate similarities against the entire vocabulary in milliseconds. The system maintains a cached ranking for each puzzle, ensuring fast, consistent responses even during peak traffic.
Every day at midnight UTC, a new target word is selected for players worldwide. Compete on the global leaderboard based on your number of guesses, time taken, and hints used. Compare your strategies with friends and see how you rank against the community.
Practice mode with no time pressure. Perfect for learning the game, exploring semantic relationships, or just relaxing with word puzzles. Take as many guesses as you need— there's no penalty for experimentation.
Real-time 1v1 competition where you race against another player to find the same target word. See your opponent's progress as you play—will you find it first? Quick matches typically last 2-5 minutes, perfect for competitive bursts. Requires a registered account.
Cooperative team mode for 2-4 players. Share your guesses in real-time, collaborate on strategy, and work together to find the target word. Perfect for playing with friends, teaching new players, or combining different thinking styles. Create or join rooms with custom codes.
Wordsy is built with a modern, scalable stack designed for real-time gameplay:
This architecture ensures sub-second response times for guesses and smooth real-time collaboration in multiplayer modes.
Wordsy is completely free to play, with all features accessible to everyone. We keep the game running through advertisements provided by Google AdSense.
Our commitment:
For details about our advertising practices and terms of use, see our Terms of Service.
Most word games test vocabulary size or pattern recognition. Wordsy is different—it explores the conceptual structure of language itself.
Because Wordsy rewards understanding over spelling, it levels the playing field:
Wordsy is more fun when you play with others. Create a free account to unlock:
Guest accounts let you try the game immediately, but registering takes just seconds and saves all your progress.
Ready to explore the semantic space and discover how words connect? Check out our How to Play guide for strategies and tips, then jump into your first puzzle!
Whether you're a word game veteran or completely new to semantic similarity games, Wordsyoffers a fresh, engaging experience that teaches you about language while you play.
Start with broad category words like "animal", "emotion", or "object". Once you see which categories rank well, narrow down with more specific words. This strategy helps you explore the semantic space efficiently!