Dex is an exploratory chess engine uniquely built in Python. The engine maintains its own gamestate and move generation, employing a minimax search complemented by several optimization techniques and a rudimentary evaluation heuristic move selection. Currenetly, Dex approximately plays at a 1400 level on Lichess in Blitz/Rapid. You can watch Dex's games or play against it when it's online!
| Feature | Benefits |
|---|---|
| Bitboard Representation | Encodes positions as 12 64-bit ints to enable efficient bitwise operations for faster computations. |
| Minimax Search w/ Alpha-Beta Pruning | Prunes unpromising move branches early in search tree, enhancing decision-making efficiency. |
| Iterative Deepening | Searches for the best move at increasing depths, leveraging previous results for optimization. |
| Quiescence Search | Addresses the horizon effect by evaluating dynamic positions more deeply. |
| Zobrist Hashing & Transposition Table | Reduces redundant calculations by caching previously evaluated positions. |
| UCI Compatibility | Allows seamless use with popular chess GUIs and automatic play via the Lichess API. |
| Cloud Computing | Offers flexibility to use local resources or serverless compute on AWS Lambda. |
If you have any questions or would like to contribute to Dex's development, feel free to fork the repo, create a pull request, log an issue, or start a discussion.