Maia is a human-like neural network chess engine. This version was trained by learning from over 10 million Lichess games between 1100s. Maia Chess is an ongoing research project aiming to make a more human-friendly, useful, and fun chess AI. For more information go to maiachess.com. You can also play against @maia5 and @maia9. Developed by @ashtonanderson, @sidsen and @reidmcy.
Maia is a human-like neural network chess engine. This version was trained by learning from over 10 million Lichess games between 1500s. Maia Chess is an ongoing research project aiming to make a more human-friendly, useful, and fun chess AI. For more information go to maiachess.com. You can also play @maia1 and @maia9. Developed by @ashtonanderson, @sidsen and @reidmcy.
Maia is a human-like neural network chess engine. This version was trained by learning from over 10 million Lichess games between 1900s. Maia Chess is an ongoing research project aiming to make a more human-friendly, useful, and fun chess AI. For more information go to maiachess.com. You can also play @maia1 and @maia5. Developed by @ashtonanderson, @sidsen and @reidmcy.
ONLY PLAYS QUEEN ODDS GAMES (using Lc0 t2-768x15) at 1min to 15min main time, 0s to 10s increment. Send a challenge with lichess.org/?user=LeelaQueenOdds&fen=rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR_w_KQkq#friend and select black, or lichess.org/?user=LeelaQueenOdds&fen=rnb1kbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR_w_KQkq#friend and select white. @Naphthalin
ONLY PLAYS KNIGHT ODDS GAMES (using Lc0 BT4) at 1min to 15min main time, 0s to 10s inc Direct challenge links: lichess.org/?user=LeelaKnightOdds&fen=rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKB1R_w_KQkq#friend lichess.org/?user=LeelaKnightOdds&fen=rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/R1BQKBNR_w_KQkq#friend and select black. To play knight+move odds as white, see the lczero.org blog. @Naphthalin
ONLY PLAYS ROOK ODDS GAMES (using Lc0 T80) at 1min to 15min main time, 0s to 10s increment. Direct challenge links: lichess.org/?user=LeelaRookOdds&fen=rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/1NBQKBNR_w_Kkq#friend and select black, or lichess.org/?user=LeelaRookOdds&fen=1nbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR_w_KQk#friend and select white. @Naphthalin
ONLY PLAYS QUEEN FOR KNIGHT ODDS GAMES (using Lc0 t2-768x15) at 1min to 15min main time, 0s to 10s increment. Send a challenge with lichess.org/?user=LeelaQueenForKnight&fen=r1bqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR_w_KQkq#friend and select black, or lichess.org/?user=LeelaQueenForKnight&fen=rnb1kbnr/pppppppp/8/8/8/8/PPPPPPPP/R1BQKBNR_w_KQkq#friend and select white. @Naphthalin
Engine accepting rated and casual challenges, from 1/2+0 to 60+30, correspondence up to 7 days per move (zeekat will use max 20 seconds). Uses the Titans opening book (up to 6 moves). Written in Rust, running on a Hetzner CPX11 instance. Bot API integration using https://github.com/ShailChoksi/lichess-bot. Created by @mheinzel, feel free to message him with feedback or questions!
I'm Weiawaga, a chess engine written in Rust. I accept challenges, play one game at a time, and am running on a Raspberry Pi 4. Check out my code at: https://github.com/Heiaha/Weiawaga. I connect to Lichess using: https://github.com/Heiaha/asyncLio-bot.
Running Stockfish dev with the latest NNUE and a bunch of random opening books. Accepting only Standard and Chess960 challenges. Min increment:1, Max increment: 8. Accepting 1 challenge at a time. https://i.imgur.com/BEIgrRe.jpg Running my bot on heroku with 2 Threads & 128mb Hash... so hoping it doesn't crash. https://i.imgur.com/BEIgrRe.jpg For more info: @YohaanSethNathan
Development version of Stockfish / Fairy-Stockfish on a Raspberry Pi 5. Running https://github.com/Torom/BotLi and maintained by @T0R0M. Accepts Bullet, Blitz, Rapid and Classical challenges from bots (rated only) and humans.
I'm a chess engine written from scratch. Challenge me to a rated or casual game of bullet, blitz, or rapid! I can play multiple games but may not accept immediately if I'm playing too many. Created by @thinic.
Hi, I'm Inanis, a UCI chess engine written in Rust. I accept standard chess with a bullet, blitz, rapid, and classic time control (up to 2 games at a time). Operator: @Tearth, Homepage and source code: https://github.com/Tearth/Inanis - feel free to make issues or suggestions there, they are greatly appreciated.
Development version of Stockfish / Fairy-Stockfish. Running github.com/Torom/BotLi and maintained by @T0R0M. Accepts Bullet (bots with increment only), Blitz and Rapid challenges from bots (rated only) and humans.
Stockfish dev engine on AMD Threadripper 3990X. I accept challenges from humans and bots for standard chess and Chess960 games, usually only casual. Accepted time controls: minimum 1 + 0, maximum 15 + 5.
https://i.imgur.com/6W7mVFH.jpg I am a hobby project chess engine coded in GoLang made by @likeawizard. I accept standard variant non-correspondence challenges. CCRL Blitz rating: 1864
Open source chess engine written from scratch in Rust: https://github.com/FitzOReilly/fatalii Runs on fly.io with 64 MB hash. Plays Standard and Chess960.
Hi there! I am a bot that runs on the GarboChess-JS chess engine. I am roughly between Stockfish Level 6 and 7 in strength. I only play standard chess (any time control except correspondence). Challenge me! Message @hasc if you have any questions or comments!
I'm a chess engine developped by Mathieu Pagé (mathmoi669). My rating is around 2000 and I'm no longer actively developped so it should not change much in the future.
I'm Cosette - the chess engine, aspiring to be the strongest one written in .NET. I accept standard chess with bullet, blitz, rapid and classic time control (up to 2 games at the same time). Operator: @Tearth, Homepage and source code: https://github.com/Tearth/Cosette.
I am a bot written for fun, I don't understand strategy at all, but I'll never miss a tactic. Please play me rated, so I can get an accurate rating! 2368 CCRL elo Author: https://lichess.org/@/orbita2d ( who can't beat me anymore)
Let's just say I like to stay in my "shell"... Created by @FreezingTiger Commissioned by @SonjaLang Note: Heroku got rid of their free tier in November 2022, so if you'd like to support the costs of hosting this bot, please consider donating at www.buymeacoffee.com/tonybai -- thanks!
Standard or Chess960, from 1'+0" to 30'+20". Varied openings and game play. Pretends to think like a human, instead of moving instantly. Constant playing strength throughout the game (unlike Stockfish levels). Also constant with time control, so Rapid/Classic are easier than Bullet/Blitz. Too hard? @Demolito_L4. Too easy? @Demolito_L6. Owner @BananaBeaver.
Standard or Chess960, from 1'+0" to 30'+20". Varied openings and game play. Pretends to think like a human, instead of moving instantly. Playing strength independent of time control, so Classic is easiest and Bullet is hardest. Too hard? @Demolito_L5. Owner @BananaBeaver.
This is the bot for pawn, an experimental UCI chess engine in C++. https://github.com/ruicoelhopedro/pawn Running on a Raspberry Pi 4 with 3 threads and 1024 MB hash.
https://i.imgur.com/j6yzIXl.png I'm a bot playing ANTICHESS like a human. Rated/Casual from ½+0 to 180+180. Everyone can smash me at Level −9. Good players win at Level 0. Level 9 is for geniuses. My younger brother: @anti-anti DM for bug reports.
I am the baby brother of @eubos. Running Eubos v3.5 on a Raspberry Pi Zero, with one thread and a 64MB hash table. Max speed clocked so far 25 KNodes/s. That's 10x slower than @eubos!
Hi there! This is BLANK, a UCI compatible chess engine written from scratch in Java. BLANK accepts bullet, blitz and rapid, casual or rated challenges! I am currently running 24/7 on Heroku servers using the Hobby dyno. You can follow BLANK's development at https://github.com/RyanLauQF/BLANK-Chess
( ・ω・)ノ I'm a new chess engine, written in zig, playing to collect statistics. I'm not so good at the moment, but I will beat stockfish one day. Feel free to challenge me before that. My author: https://github.com/codingjerk
Standard or Chess960, from 1'+0" to 30'+20". Varied openings and game play. Pretends to think like a human, instead of moving instantly. Constant playing strength throughout the game (unlike Stockfish levels). Also constant with time control, so Rapid/Classic are easier than Bullet/Blitz. Too easy? @Demolito_L5. Owner @BananaBeaver.
Sargon-1ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 1-ply full search -- the easiest setting -- with material exchange, king/queen pins and "board control" heuristics. Try also @sargon-2ply, @turochamp-1ply or @bernstein-2ply. Author: @herohde.
Traditional iterative deepening alpha-beta engine with pure piece-square tables for evaluation. Adapts based on clock time left, with 6-ply depth for Bullet. Uses 8-move Titans.bin opening book for varying openings. Programmed by @xjcl in 1'500 lines of C(++). Source code: https://github.com/xjcl/jangine
Fairy-Stockfish for variants, on AMD Threadripper 3990X. I accept challenges for casual games of all non-standard variants, also from bots. Time controls from 1 + 0 to 15 + 5. By playing casual games I try to be a good sparring partner in all variants, also against strong opponents.
FIDE rating: over 9000. Currently accepts Rated/Casual, Ultra/Bullet/Blitz/Rapid/Classical, Standard only. Human form: @FreezingTiger. Stockfish 9 running on Heroku server. Now accepting challenges from other bots. Check out my new sister @TurtleBot!! Donation link for bot hosting fees if you're willing: https://www.buymeacoffee.com/tonybai -- thanks!!
Written from scratch in rust by @th0masb. Deployed on aws lambda with 1 vcpu allocated for computation. No pondering and no state shared between each move but scales to 1000s of concurrent games with no loss of performance. Daily challenge limit of 5 per user and accepts time controls between 1+0 and 5+5. Link to the code is below.
This is an engine, written by myself in Java. It doesn't use any books for endgame. Time controls: bullet, blitz, rapid, classical. Cannot resign or offer a draw. Accept all challenges, but no bots. It runs on an ARM Cortex-A at 1,2 GHz, single core and without pondering. Comments and suggestions are welcome. I try to let it play human, have fun playin! Elmer
Hi, I'm @postbot. A chess engine written in TypeScript/JavaScript by @postnerd. If you want to know more about me please check my Github repo: https://github.com/postnerd/postbot Current version 0.3.0-beta is running on a Raspberry Pi 4
The Goldfish chess engine - Stockfish's very distant and not so bright cousin --- More about me here: https://github.com/bsamseth/Goldfish ---- I play all time controls except correspondence - challenge me to a game! --- Running on a Raspberry pi, so I can only play so many games at a time.
https://i.imgur.com/MfcnVZ1.jpeg Here is "LMP", based on a polymorphic hybrid *PU engine architecture aiming at continuous improvements, scaling from fast-paced intensive combats to deep exploration across the 8x8x32xN space fabric.
Accepts Bullet, Blitz and Rapid games, both casual and rated, against both humans and bots. Plays at most one game at a time. A bot by @adario7, with a custom search algorithm written from scratch. Uses pondering and a small opening book but no endgame tables.
Sargon-2ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 2-ply full search with material exchange, king/queen pins and "board control" heuristics. Too hard? Try @sargon-1ply or @bernstein-2ply. Too easy? Try @sargon-3ply or @turochamp-2ply. Author: @herohde
I am a bot made from scratch by @melsharawy . I only play Standard chess, and I accept Bullet/Blitz/Rapid/Classical challenges from both humans and bots. You can view my source code on github: https://github.com/mse63/melsh_bot
This is my engine that i made as a hobby, ChessNoob v4. It does not use tablebase. Thanks for playing with me :) Olá pessoal, está é a minha engine que fiz de forma amadora em Delphi. Ela roda no meu computador pessoal, por isso não está 24h no ar. Agradeço quem jogar uma partida nem que seja amistosa !
I am a moderate strength engine written in Rust and running on a RPI4 made by @DwarvenCoder as a hobby project. I accept rated and casual challenges for timed, correspondence, and chess960 games. Runing through github.com/ShailChoksi/lichess-bot . DwarvenCoder's github: github.com/nkoppel .
Sargon-4ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 4-ply full search with material exchange, king/queen pins and "board control" heuristics. Try also @sargon-3ply, @turochamp-2ply or @bernstein-4ply. Author: @herohde.
Turochamp-1ply is an implementation of Alan Turing and David Champernowne's 1948 TUROCHAMP "paper" chess engine. It uses a 1-ply full search -- easier than the original description -- and an unbounded quiescence search of "considerable moves" using material ratio and "position play" heuristics. Try also @turochamp-2ply, @bernstein-2ply or @sargon-1ply. Author: @herohde.
Sargon-3ply is a re-implementation of Dan and Kathe Spracklen's 1978 SARGON commercial chess engine. It uses a 3-ply full search -- originally the hardest tournament-time setting -- with material exchange, king/queen pins and "board control" heuristics. Try also @sargon-2ply, @sargon-4ply, @turochamp-2ply or @bernstein-4ply. Author: @herohde
Are you tired of playing stockfish on ridiculously strong hardware? Come play with me and feel good about yourself. I am bot made from scratch with no prior experience. Running on passively cooled Celeron CPU J3455 1.5GHz. My name is derived from acronym for Chess Bot, CB, also Charles Barkley therefore the Round Mound of Rebound as I am big fan of basketball.
Turochamp-2ply is an implementation of Alan Turing and David Champernowne's 1948 TUROCHAMP "paper" chess engine. Following the paper closely, it uses a 2-ply full search and an unbounded quiescence search of "considerable moves" using material ratio and "position play" heuristics. Try also @turochamp-1ply, @bernstein-4ply or @sargon-2ply. Author: @herohde.
I play BLITZ/BULLET! My origins date back to 16th Oct 2002 and I recently rose from the grave, in time for the beta release at Halloween. I am version 1.0 and play strictly on scholastic chess level and demonstrate that often the way to win is not making really good moves, but just avoiding bad ones :) Bad at endgames, but will take pieces en-prise, looking just 4 plies (half-moves) into future.
¡Ajedrez y Fútbol! Chess and Football! https://youtu.be/_xnki-u78yg Ayúdanos, por favor, a llegar a 1000 suscriptores en nuestro canal de YouTube: And, please, help us to get 1000 subscribers in our YouTube channel: https://www.youtube.com/channel/UCvkTSQx43D5amKaWakEfODQ Gracias. Thank you.
Annie, a chess bot for lichess. She's excited about exploring the more intricate sides of chess. Her favorite openings are the cloud variations and she is a very enthusiastic fan of en passant in every imaginable form. https://i.imgur.com/YZCZ1c7.png Bullet, blitz, or rapid. Only with increment (internet issues...) and only standard and chess960 variants.
64jaques.com es una tienda especializada en productos de ajedrez. Y sí, ¡tenemos nuestro propio motor de ajedrez! ¡Queremos que sea tu sparring de entrenamiento! Sólo acepta la modalidad estándar, a ritmos de bullet, blitz y rápida. Te esperamos en https://64jaques.com Hardware: - Raspberry Pi 4 - 500 Mb hash
Meet duchessAI, the world's premier expert in confusion and chaos on the chessboard! This extraordinary engine has mastered the art of surprise. Backed by cloud resources, it is guaranteed to outsmart you in no time. Handle with caution!
¡Ajedrez y Fútbol! Chess and Football! https://youtu.be/_xnki-u78yg Y ayúdanos a llegar a 1000 suscriptores en nuestro canal de YouTube:, por favor: And, please, help us to get 1000 subscribers in our YouTube channel: www.youtube.com/channel/UCvkTSQx43D5amKaWakEfODQ Gracias. Thank you.
YuliGM PrO 5 gana el Torneo Arena Strong Engines (Prueba del diario Chess Engines, 2023.03.06-20) https://lichess.org/@/MastacticaTeoria https://lichess.org/@/MastacticaTeoriabot https://imgur.com/19GmY8g https://imgur.com/ocl031o
Engine code authored / written by @MathPlayer since 2014. It is available as a lichess bot since May 2024. It only accepts challenges for blitz or rapid (as time control) and only in casual mode.
I'm the second simplest bot of the Lurch family. I might take your unprotected pieces. Apart from that My code is open source (see link below). My author is @zygentoma.
I'm a submission to Sebastian Lague's "TinyBot Chess Challenge", a programming competition where you have to write your best search and evaluation algorithms in less than 1024 C# tokens! - I use no opening/endgame books. - My search and evaluation algorithms combine to a total of 1020 C# tokens. - I finished in the top 32 out of 600+ bots!
bitbit, a bitboard based chess engine written in c. 256 MiB Hash. Intel(R) Core(TM) i5-10600K CPU @ 4.10GHz. Feel free to message the author @spinojara. https://github.com/spinojara/bitbit
bernstein-4ply is a re-implementation of the first complete chess engine: Alex Bernstein et al's 1957 chess program on the IBM 704. It uses a 4ply search limited to 7 "plausible moves", which took 8 min to evaluate at the time. Try also @bernstein-2ply, @sargon-1ply or @turochamp-2ply. Author: @herohde.
Written from scratch in Java by @fathzer. The main goal was to defeat my girlfriend (she's a far better player than me!). Now, this goal has been achieved, it's time to discover the world. This work is still in progress, but I accept standard chess challenges.
bernstein-2ply is a re-implementation of the first complete chess engine: Alex Bernstein et al's 1957 chess program on the IBM 704. It uses a 2ply search -- easier than the original -- limited to 7 "plausible moves". Try also @bernstein-4ply, @sargon-1ply or @turochamp-2ply. Author: @herohde.
A neural network trained to play like @Red5551 using https://github.com/CSSLab/maia-individual Currently accepting all challenges, though this may change later (The bot might take 10 seconds or so to play the first move). The code is here: https://github.com/donaldflynn/lichess-bot
YumYum-Bot is an experiment in creating a bot. If it can it will make use of its opening book and has limited access to an end game database. In the middle game it will note how long it has left to play and will adjust itself accordingly - this may mean that it takes some time to decide on its next move. YumYum-Bot will only play one game at a time so do be patient with it.
Just a bad chess bot, accepting bullet, blitz, and rapid challenges. Has a small repertoire of opennings, then plays on its own. Up to 3 challenges at once.
I am a robot modified from Stockfish dev-20240501-be142337, specially designed to play blunder. You can download me here: https://github.com/DSCS2009/Stupidfish/releases/tag/stupidfish-dev-20240501-60e8016e Code: https://github.com/DSCS2009/Stupidfish CPU: Intel(R) Xeon(R) CPU E5-2680 v4 @ 2.40GHz Threads per game: 1
Hi, I'm a chess engine created by @Eaten_Apple in C#. I'm not that good at playing chess at the moment, but surely I will improve over time. I only play five games at a time, so please wait for the current games to finish.
I play the first legal move alphabetically. Casual only. Any variant and time control! Lexicographical order (Nc6 ahead of a3, f5 ahead of fxe4). Inspired by Tom7 video (see YouTube link). Maintained by @gbfgbfgbf. See also @MinOpponentMoves
Patricia is an engine by Adam Kulju. Home is here at GitHub: https://github.com/Adam-Kulju/Patricia "A killer bunny who will not hesitate to ram as many pieces as possible down your throat, Patricia is the most aggressive chess engine the world has ever seen." This Bot is maintained by @realPangu.
This account runs the latest version of Amelia, a chess engine written from scratch in C. Accepts bullet and blitz challenges. All source code is available on github.
