Intelligent Wordle assistant powered by information theory with 12,972+ word database
A fast and efficient solver that provides optimal word suggestions to help you win at Wordle using entropy-based calculations.
- Information Theory Algorithm - Optimal suggestions based on entropy calculations
- Comprehensive Database - 12,972 valid five-letter words
- Efficient Input - Quick feedback entry using G/Y/B notation
- Keyboard Shortcuts - Tab for autocomplete, Enter to process
- Clean Interface - Modern, professional design
- Undo Function - Correct mistakes without restarting
- Live Statistics - Track attempts and remaining possibilities
-
Clone the repository:
git clone https://github.com/sersaumy/wordle-solver.gitcd wordle-solver -
Run the application:
python main.py
No additional dependencies required. Uses built-in Tkinter.
- Launch the application - Run
python main.py - View suggestion - The solver displays the optimal word (e.g., "SOARE")
- Play Wordle - Enter that word in your Wordle game
- Enter feedback:
- Type your word (or press Tab to auto-fill suggestion)
- Enter feedback pattern:
- G = Green (correct position)
- Y = Yellow (wrong position)
- B = Black/Gray (not in word)
- Example: GYBBB means Green, Yellow, Black, Black, Black
- Get next suggestion - Repeat until solved
| Key | Action |
|---|---|
| Tab | Auto-fill suggested word |
| Enter | Move to next field / Process guess |
| Mouse Wheel | Scroll through interface |
Attempt 1: SOARE -> Feedback: BBYBG Attempt 2: BLUNT -> Feedback: BGYBG Attempt 3: FLINT -> Feedback: GGGGG (Solved)
- Tab autocomplete for instant word filling
- Color-coded attempt history
- Real-time statistics
- Undo last attempt
- Scrollable interface
The solver uses information theory to maximize information gain with each guess:
- Word Bank - Loads official Wordle answers (2,309) + valid guesses (10,663)
- Entropy Calculation - Scores words based on expected information gain
- Constraint Filtering - Eliminates impossible words based on feedback
- Optimal Selection - Suggests words that narrow down possibilities fastest
- Entropy-based word scoring
- Green letter position matching
- Yellow letter inclusion with position constraints
- Gray letter elimination
- Duplicate letter handling
- Average solve rate of 3.4 attempts
- Language: Python 3.7+
- GUI Framework: Tkinter (built-in)
- Data Source: Official NYT Wordle word lists
- Algorithm: Information theory with constraint propagation
| Metric | Value |
|---|---|
| Success Rate | 100% within 6 attempts |
| Average Attempts | 3.4 |
| Best Case | 2 attempts |
| Word Database | 12,972 words |
Contributions are welcome. Please feel free to:
- Report bugs
- Suggest features
- Submit pull requests
MIT License - Free to use for personal or educational purposes.
Live Demo: https://sersaumy.github.io/wordle/
No installation needed! Use it directly in your browser.