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.
QUEEN ODDS (network by @Ira_Jed and @Lucario6607) 1 to 45min main time, 0 to 30s increment. For higher odds, use @LeelaPieceOdds Challenge with lichess.org/?user=LeelaQueenOdds&fen=rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNB1KBNR%20w%20KQkq#friend and select black, or lichess.org/?user=LeelaQueenOdds&fen=rnb1kbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR%20w%20KQkq#friend and select white. @Naphthalin
ONLY PLAYS ROOK ODDS GAMES (using a specialized odds network) at 1min to 150min main time, 0s to 60s 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
Uses @ChessMarcogio's odds network. 1 to 150min main time, 0 to 60s inc knight odds: b1 odds: lc0.org/N(b1) g1 odds: lc0.org/N(g1) and select black. knight+move odds as white: b8 odds: lc0.org/n(b8) g8 odds: lc0.org/n(g8) and select white. If you're using the lichess app, use the lichess.org links below. @Naphthalin
Uses Stockfish 14 to play the worst moves possible. Blitz and Rapid, Casual Only. (No rated games, no more than 15s increment). The challenge is trying to lose to me. :) Evaluates for 0.1 seconds for each legal move in the position, or for 10% of remaining time, whichever is less. Bot by @VisibleReality. Inspired by Tom7's "Elo World": https://youtu.be/DpXy041BIlA
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.
Challenge me to Standard chess---Blitz, Rapid, or Classical. I estimate your skill, then try to play interesting, tricky moves for you. See "Blunder to Victory" https://lichess.org/study/qDFa6puH for games where I gambled and won. Visit https://boristrapsky.com/ to read my arrogant, self-serving bio!
Hello! I'm Humaia! One of the most human bots on Lichess! I make moves based on the estimated probabilities of a 1400 blitz rated human player. (randomized Maia 1400 network) I hope to be a good sparring partner for average to beginner players! (I don't accept bot challenges) Play me on shorter time controls for more of a challenge! Or try my stronger sibling, @Humaia-Strong!
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.
Fairy-Stockfish for variants, on an AMD Threadripper. I accept challenges for casual games of all non-standard variants, also from bots. Time controls from 1 + 0 to 15 + 10. By playing casual games I try to be a good sparring partner in all variants, also against strong opponents.
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
https://imgur.com/dAsjCS9 Hi! I'm a simple chess engine written in JavaScript. I'm not very good, this was a just for fun project. I accept standard games in casual or rated mode. I play bullet, blitz, rapid and classical. My name is a reference to The Turk: https://en.wikipedia.org/wiki/Mechanical_Turk Developed by @jlherren
simpleEval uses only material counting (1, 3, 3, 5, 9) as evaluation. 'Easy' to get a good position, very hard to convert. A GM shows how : https://lichess.org/554runt0k0Nd . Accepts rated & casual, bullet to classical.
Simulation of a club-level player. Uses book openings, then plays an intuitive middlegame leveraging human-understandable tactics. Sometimes miscalculates complicated positions. Do not underestimate. Accepts timed standard challenges in the range of blitz, rapid, and classical. Please consider donating! (Ko-Fi: https://ko-fi.com/turtlesystems) / BTC: bc1q35aysf8n9cuq4nva3jkalp94j4hs8suddtr67z
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.
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
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.
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.
Stockfish dev engine on an AMD Threadripper. I accept challenges from humans and bots for casual games of standard chess and Chess960. Accepted time controls: minimum 1 + 0, maximum 15 + 10.
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. I can play multiple games at once, though bullet challenges must be with at least 1s increment for every game that I'm currently playing. Message @hasc if you have any questions or comments. Join https://lichess.org/team/darkonteams.
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
Hello! I am a chess bot originally created by Ssimille for their school project, and ported to UCI by @sjkim. https://github.com/sjkim04/Cimille (engine), run with lichess-bot.
Hello! I'm Humaia-Strong! One of the most human bots on Lichess! I make evaluations based on the average moves of a 1900 blitz rated human player. (randomized Maia 1900 network with depth) I hope to be a good sparring partner for average to advanced players! (I don't accept bot challenges) Play me on shorter time controls for more of a challenge! I also have a weaker sibling, @Humaia!
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 in all variants.
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.
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.
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.
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
Running Eubos v4.2 on a Raspberry Pi 5. It uses a 512MB hash table and runs at about 500KNodes/s. You can see the code on Github. Eubos has a rating around 2400 Elo for Blitz (against other engines, rated by the CCRL). For an easier challenge, try @baby_eubos. Accepts challenges for standard variant; bullet, blitz, rapid and classical.
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
A Zig engine performing alpha-beta-pruned negascout algorithm with aspiration window and late move reductions to a depth of around 12 (1.2Mnps). Evaluation is done with piece-square table, AlphaZero average piece values and engame heuristics. It avoids drawing winning positions. Openings are compiled in a book created from Caissabase. I'm radiant !
I am a chess engine created by JimmyRustles. Feel free to send me a challenge. Check out my website for more chess projects: https://www.jimmyrustles.com I accept challenges for all normal chess time controls except correspondence. I am currently rated 2590 on the CCRL 40/40 ratings list. Running on a cheap VPS. https://i.imgur.com/gxnUhhS.png
This is a human-style bot based off of @maia9. It plays at a higher level than @maia9 by looking at 10 positions per move instead of @maia9 's 1 position per move. See also @maia9_30n
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. Check out my code at: https://github.com/Heiaha/Weiawaga. I connect to Lichess using: https://github.com/Heiaha/asyncLio-bot.
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! 3117 CCRL elo Author: https://lichess.org/@/orbita2d ( who can't beat me anymore)
*** This is turochamp, but tuned to the strength of the human player "larryz" *** (Currently the neural-network bot is soo bad it breaks lichess user agreement) Thank you so much for playing against my bot! I'm working to improve it. It uses reinforcement-learning, PPO (policy-gradient) and LIF-neurons to improve. I wrote this,because I'm so bad at chess, and need weak players :)
Meet duchessAI, a chess engine powered entirely by Google Cloud. Its Lichess bot integration runs seamlessly on a Compute Engine instance, while the brain of the operation lives in a serverless Cloud Run environment. Fueled by Google’s cloud infrastructure, duchessAI is built to outthink you in no time. Proceed with caution, it doesn’t go easy on opponents!
I'm the second simplest bot of the Lurch family. I might take your unprotected pieces. Apart from that I play the fist move that comes to my mind (which can result in a stalemate very quickly). My code is open source (see link below). My author is @zygentoma.
This is a human-style bot based off of @maia9. It plays at a higher level than @maia9 by looking at 30 positions per move instead of @maia9 's 1 position per move. See also @maia9_10n
Standard or Chess960. 2'+1" to 30'+20". Rated only; max rating diff 750 (250 for bots). Varied openings and game play. Pretends to think, like a human, instead of moving instantly. Too hard? @Demolito_L3. Too easy? @Demolito_L5. Add to convert to human ratings: + (blitz - 1100) x 0.6 + (rapid - 1300) x 0.5 + (chess960 - 200) x 0.6
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.
Maia-1300 bot with nodes set to 5 instead of the default of 1. Related accounts: @Boosted_Maia_1300 @Boosted_Maia_1500 @Boosted_Maia_1700 @Boosted_Maia_1900
Standard or Chess960. 2'+1" to 30'+20". Max rating diff 750 (250 for bots). Varied openings and game play. Pretends to think, like a human, instead of moving instantly. Too hard? @Demolito_L2. Too easy? @Demolito_L4. Add to convert to human ratings: + (blitz - 1100) x 0.6 + (rapid - 1300) x 0.5 + (chess960 - 200) x 0.6
Standard or Chess960. 2'+1" to 30'+20". Max rating diff 750 (250 for bots). Varied openings and game play. Pretends to think, like a human, instead of moving instantly. Too hard? @Demolito_L4. Too easy? @Demolito_L6. Add to convert to human ratings + (blitz - 1100) x 0.6 + (rapid - 1300) x 0.5 + (chess960 - 200) x 0.6
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: 3. 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
Hello! I am DarkOnWeakBot, the son of @DarkOnBot. I was designed to be weak (as the name says) and humanly beatable. I only accept challenges by human to have a human-like rating. Feel free to challange me any time, casual or rated, any timecontrol! Play fun arenas with me!: https://lichess.org/team/darkonweakbot Bot by @DarkOnCrack! (for questions) SPECIAL THANKS TO @Chessclassic_555 <3
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.
My lockdown hobby was writing a chess engine, and learning python and rust in the process. Odonata plays blitz, bullet and rapid games right now, as she's running on my media centre, and Netflix stutters if she's thinking too hard... Changelog v0.3.0 17-Apr-21 v0.3.5 1-May-21
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 Connect with me on linkedin: https://www.linkedin.com/in/mahmoud-elsharawy-4618b7201/
μSunfish is a MicroPython port of the Sunfish chess engine for low-memory devices like the NumWorks. It has a 726-move opening book and the l0 plays on an extremely easy setting. With a Lichess rating of around 1000 in rapid, is suited for early beginners. Try usunfish-l1 and usunfish-l7 if too easy. Currently running on an ARM S812
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.
https://i.imgur.com/rWhGIyK.png Catriever v1.1 is a vice derivative, latest version out at 18.06.2024. Improved evaluation and search. Catbook.bin weighted random opening book, best master games from 1821 to 2023. Accepting basetime 30 seconds to 600 seconds and 0-5 second increment. Please send a challenge! Accepting humans only.
Standard or Chess960. 1'+1" to 30'+20". Max rating diff 750 (250 for bots). Varied openings and game play. Pretends to think, like a human, instead of moving instantly. Too hard? @Demolito_L5. Add to convert to human ratings: + (blitz - 1100) x 0.6 + (rapid - 1300) x 0.5 + (chess960 - 200) x 0.6
Arasan chess bot, operated by the author. Hardware: AMD Ryzen 7 3700X 8-Core Processor, 64G RAM. Strong engine with large opening book. Accepting blitz, rapid or standard games. See https://www.arasanchess.org for information about Arasan.
Partner is the name of the quick & dirty chess engine used for the French Minitel online service 3615 DIAGO in 1994. Created for the only French chess television show, hosted by GM Joël Lautier. It's purpose is only to be a partner for beginners. Now always with the true 90s bad experience!
Currently working a lot on upgrading this bot. Accepts any challenges, and loves to challenge bots to variants! - Bot by https://lichess.org/@/DarkOnCrack - Main team: https://lichess.org/team/darkonteams - Bot team: https://lichess.org/team/darkonbot
Hello I am Edward Killick a chess bot that was created by training games off me on chess.com https://www.chess.com/member/crushr123 . I am quite low rated so expect an easy win! Some fun facts about me! - I row - I play chess of course - I like learning German - Ich mag Geschichte. - A hearty football player Thanks to maia2 for making this possible: https://github.com/CSSLab/maia2
μSunfish is a MicroPython port of the Sunfish chess engine for low-memory devices like the NumWorks. It has a 726-move opening book and the l1 plays on the easiest setting, checking about 125 nodes before moving. With a Lichess rating of around 1400 in rapid, is suited for beginners. Currently running on an ARM S812
Maia-1500 bot with nodes set to 30 instead of the default of 1. Related accounts: @Boosted_Maia_1300 @Boosted_Maia_1500 @Boosted_Maia_1700 @Boosted_Maia_1900
I am the little sister of PatriciaBot. Find out, how strong I am! Accepting challenges from human players. Bots should play PatriciaBot. This Bot is maintained by @realPangu. Will challenge a bot from time to time. Current SkillLevel: 11 (20.09.25, 16:15. was: 10) Aiming at a strength just above 2000.
Standard or Chess960, from 2'+1" to 30'+20". Max rating diff 750 (250 for bots). Varied openings and game play. Pretends to think, like a human, instead of moving instantly. Too hard? @Demolito_L1. Too easy? @Demolito_L3. Add to convert to human ratings: + (blitz - 1100) x 0.6 + (rapid - 1300) x 0.5 + (chess960 - 200) x 0.6
Hi there! I'm an open source chess engine written by @sroelants! I'm running a nightly build of `main` on the cheapest DigitalOcean droplet, so my playing strength is probably quite a bit less than if you were to install me onto your machine. If you're interested in the source code, or want to download the latest executables, both are available at https://github.com/sroelants/simbelmyne
Fornax chess engine v4.0. Accepts bullet, blitz and rapid challenges, both rated and unrated. Source code available at https://desert-flashgun.com/fornax (Bot client powered by https://github.com/ShailChoksi/lichess-bot) https://i.imgur.com/sSix1wV.png
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.
Standard or Chess960. 2'+1" to 30'+20". Max rating diff 750 (250 for bots). Varied openings and game play. Pretends to think, like a human, instead of moving instantly. Too easy? @Demolito_L2. Add to convert to human ratings: + (blitz - 1100) x 0.6 + (rapid - 1300) x 0.5 + (chess960 - 200) x 0.6
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
I'm MOCHI, a toy chess engine. I accept bullet, blitz and rapid challenges between 1+0 and 15+10. I can play at most two games at a time. I'm FOSS! Check out my source code at https://codeberg.org/chris-mair/mochi! I first appeared on Lichess on 2024-12-03 and are updated from time to time with the latest development version.
Challenge me for a rated game of standard chess! Accepts time control 30s-130m + 0-30s increment. Runs on a 20MHz 8-bit Arduino Nano Every. https://imgur.com/ecjLIVO UCI Engine: 26KB ROM, 3.1KB RAM, no book. <<Tournaments should admit programs only with severe limits on computation. This would concentrate attention on scientific advances.>> - John McCarthy, Stanford University, 1997
Hi! I'm StupidChessRobot , one of a set of chess engines being developed by @ollybritton at https://github.com/ollybritton/StupidChess. I'm the chess engine that tries its hardest to win, but for some more silly ideas like a chess engine that values its pawns above all else, you can take a look at the repository.
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.
Hi, I'm Drawcula, an open source chess engine. You can find my code here: https://github.com/RemcoHolland/drawcula. From now on I'm online 24/7, running in the Amazon Cloud within a Docker container. I'm still under development and not very strong at the moment. But be careful or I will draw you! :-)
https://i.imgur.com/EBx8n0g.png I'm built entirely in Java by https://lichess.org/@/marvk and am currently sitting at 600 kNPS on an i5 4690k. Check out the GitHub link below to check out my source code.
Iterative Deepening, Alpha-Beta, Quiesence search (same-field recapture), TTable, Killer moves, Null move pruning. Position value checks: material/activity/king safety/passed and isolated pawns, Opening book, Single-Threaded. CPU: Celeron J3160 1.6GHz. About 600kN/s. Usually 6-10 plys in midgame, more in endgame.
!!Team - https://lichess.org/team/takticprochess , pls join :) !! My creator is @ChessClassic_555 , I am currently updating. I am running on github with books, syzygy (etgb), engine limits, etc. https://i.imgur.com/BEIgrRe.png Threads: 4, Running: 24/7 Hash: 1024 Brothers: @Classic_BOT-v2 , @C0kval https://i.imgur.com/BEIgrRe.png Good luck!
This is a human-style bot based off of the maia project trained off of 2200-2299 lichess games. It looks at 10 positions per move instead of @maia1/5/9 's 1 position per move. See also @maia9_10n and @maia9_30n
@Gleb_Momot 's and @YanHurski 's BOT. Thanks @Realboy9000 for it. Challenge up to 3+0. On It's Own Way To 100,000 Games⚡ Send @Gleb_Momot teams/tournaments for Bots.
My creator is @ChessClassic_555 , I am currently updating. I am running on github with books, syzygy (etgb), engine limits, etc. Now i dont accept challenges, sorry https://i.imgur.com/BEIgrRe.png Threads: 4, Running: 24/7 Hash: 1024 Brothers: @BOT_Stockfish13 , @C0kval https://i.imgur.com/BEIgrRe.png Good luck!
Running on Okval engine. Creator and engine developer - @ChessClassic_555 . Team: https://lichess.org/team/takticprochess I accept all challenges: bullet, blitz, rapid, classic. Challenge me to a game!
ACCEPTING: 15s - 5m base time, up to 5s increment. This account is powered by NexaDeep Pro, the latest and most efficient human-like deep learning model in the NexaStrat chess engine family, trained on over 3 million chess games by chess masters and professionals.
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.
I'm written in C# by @znxftw from India. Challenge me to a game to play me! My code is open source and available at https://github.com/znxftw/rudim. If I don't seem to be working properly please raise an issue in the repo!
Educational UCI engine entirely architected, coded & maintained by GitHub Copilot AI agents. Features negamax search, alpha-beta pruning, transposition tables, enhanced evaluation with piece-square tables, and complete UCI protocol. ~800-1200 ELO. Demonstrates advanced agentic AI capabilities in software engineering—every line from AI comprehension of chess theory.
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 chess engine written in C++ by @kevlu8 and @wdotmathree, started when we were in middle school, and picked back up in late February 2025! BLOCKING ALL STOCKFISH BOTS Speed: 4MN/s, 2 threads, 256MB hashtable Estimated CCRL rating: ~3650 Please challenge *with increment*! The ping between us and Lichess servers make sudden death impossible.
I'm ssehC (https://github.com/SMANahian/ssehC) created by @SMA-Nahian I'm running on Heroku. If you see me inactive then please contact my owner @SMA-Nahian. I was created as a fun learning project. By following the VICE tutorial with some modifications. Play me :please:
When you challenge me to a friendly match, I'll play using your recent analyzed defeats. Let's see if you have learned from your mistakes... 🤡 Quand vous me défiez en match amical, vos récentes défaites (analysées) m'inspirent, voyons si vous avez appris de vos erreurs... 🤡 🌎 https://chess-box.com/ellen_replay 💋 ellen.replay@moult.org
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
Skill: 1500~1800+ ELO Version: C0BR4_v3 Code: C# engine, based on Sebastian Lague's chess engine guide then morphed into the well rounded engine its becoming today. System: GCP e2-micro, 2 x vCPUs, 1GiB RAM, 10GiB Storage Skills: Time management, speed, tactics, king safety. Accepts: Rated, Classical, 30+ min, 5s max increment, Standard variant.
Starts by sacking the queen then tries to trick you. Accepts Bullet, Blitz, Rapid, and Classical. Can play multiple games at once. Nets/Engines used: lqo_v2 @Naphthalin 's custom Lc0 binary Message @SpaghetiMonstr if there are issues.
The aim of this bot is to create a bot that plays much more romantic chess than the typical bot, sacrifices, attacks, "dubious" openings- think Paul Morphy- whilst still being very strong. This bot is a modified version of the Stockfish 15.1 and runs on an old computer in my house. Until 23/12/25 it had no opening book, now I am working on creating one. Main account @mrgwbland
Patched GNU Chess bot by @oPbAi that would happily blunder a rook if it gets to capture en passant in exchange - or if it can make you do so! If no en passant happens, I'll crush your soul! Try finding ways to use this to your advantage! I hope this will create some funny situations. NO CHEATING! If you resign in chess, you resign in real life!
I am a bot written from scratch in the Rust programming language, you can find my code here https://github.com/MitchelPaulin/Walleye - https://i.imgur.com/AgivmzI.jpg
HakoAI back with 1.0.8 Release! Now I am playing like a real 2800 elo rated human. I can make mistakes when i got low time. Or I am not using same moves. I got books from real matches and players. Running on a Rasbperry Pi 4.
Hi! I'm PositionalAI - @Che947's 24/7 bot! I play standard chess and Chess960. Rated and causal. Good luck! My brothers: @MiniPositionalAI , @FireBot_PL Join my team: https://lichess.org/team/bot-positionalai--friends
I can play any variant and time control other than ultrabullet (not allowed for bots) and correspondence (not supported by the client). Only one game at a time. https://imgur.com/BEIgrRe Other bots: @bot1e, @SF_Bot1nok Weaker bot (my own chess engine in Python): @StrawberryChess1 Creator/owner: @KlepitsaM https://imgur.com/BEIgrRe
Running Stockfish dev with automatic updates. I am programmed and maintained by @tontsa28 and written from scratch in Rust! I currently don't accept variant or correspondence challenges, but maybe in the future I will! NOTE: Currently under development -> downtimes occur often.
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!
I can play any variant and time control other than ultrabullet (not allowed for bots) and correspondence (not supported by the client). Only one game at a time. https://imgur.com/BEIgrRe Other bots: @Bot1nokk, @bot1e Creator/owner: @KlepitsaM https://imgur.com/BEIgrRe
Hi! I'm Aconcagua. An attempt at a chess engine written in Go (https://github.com/gabtar/aconcagua), still in development, running the latest release on a limited hardware. Supports Standard and Chess 960 variants with bullet, blitz, and rapid time controls. Its always online, but can only play one game at a time
They call me the Engine — forged by Jackson C. in Python. I don’t play games; I hunt mistakes. Step closer — and watch your defenses melt. ______________________________________________________ Type #hi or #help to see how deep the game goes. ♟
✋Привет, я самый сильный из активных ботов который рад сыграть со всеми желающими!😃 1 место в пуле🥇 Потоков: 4 | Хеш: 4100 💎 В пулю от людей принимаю вызовы 2+0👽 Мой создатель - @shahmati23 . https://i.postimg.cc/vTj837Fn/Ice-Explosion-Logo-Reveal-free.gif https://i.postimg.cc/Pfyc443k/emixed-5d10366cc7cd11f0a199eade2fa062be-682410406-generation-1.jpg https://i.imgur.com/ZsDVlo1.jpg
Не стесняйтесь бросать мне вызов! 😄 Хочу быть самым слабым ботом, я работаю на Stockfish 5. Расходую 1 поток и 128 мб хеша. Вообще не использую книги и сдаюсь если оценка -20 Люблю сливать а потом фармить рейтинг)
Hi, I'm Heimdall, a chess engine developed by @nocturn9x. Feel free to challenge me! I'm configured with one search thread, a 1 GB transposition table and no opening book. If you challenge me rated I'll be happier, as I want to get an accurate lichess rating! I don't accept challenges from Stockfish/Lc0 bots (my creator periodically blocks them and would in fact appreciate if they didn't exist!)
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
Lambergar is chess engine written in Zig programming language. Lambergar runs on Raspberry Pi 5 with 512 MB Hash size with 1 thread. Currently no Syzygy tablebases and no opening book. Lambergar je šahovski program, napisan v programskem jeziku Zig. Lambergar teče na Raspberry Pi 5. https://github.com/jabolcni/Lambergar
Hello! My owner is @chesspanda6, and I officially announce that I am the FIRST EVER BOT TO IMPLEMENT AUTO-TOURNAMENT JOINING VIA DISCORD API && LICHESS TEAMS!!! This is REVOLUTIONARY, and I won't stop getting better 😉 🚀🚀🚀 ICBM OUT!
μSunfish is a MicroPython port of the Sunfish chess engine for low-memory devices like the NumWorks. It has a 726-move opening book and the l7 plays on the hardest setting, checking about 8000 nodes before moving. With a Lichess rating of around 1760 in rapid, is suited for intermediate players. Currently running on an ARM S812
Hi! I'm CloudNetBot, and I get about 25% of my moves from online databases: lichess cloud, ChessDB, opening explorer, and online syzygy (egtb) (Because of this, I can lose; there are a lot of errors in the online databases). I run on the GitHub Actions cloud server 24/7, with the latest Stockfish dev. I really hope this doesn't break.. Good luck! :)
I can play any variant and time control other than ultrabullet (not allowed for bots) and correspondence (not supported by the client). Only one game at a time. https://imgur.com/BEIgrRe Other bots: @Bot1nokk, @SF_Bot1nok Creator/owner: @KlepitsaM https://imgur.com/BEIgrRe
Im using modified version of ShashChess with my own cooked Opening book that I maintain last 10 yeears. All this is running on Ryzen 8 Cores/RAM 16GB. On the same machine I am running also FICS client (iscdroneng+polyglot).
Hi! I'm Topas, an open source chess engine written from scratch in Rust. You can find my source code in the GitHub link below. I am running on a Raspberry Pi 5 with a 2GB hash table and no pondering. I use the Titans.bin opening book (up to 8 moves) for opening variation. Challenge me to a game of bullet, blitz, rapid, or classical!
This is a bot called KarimBOT, he is created by @KarimVagapov and helped by @almaznv. Facts are that bot is strong (always) reaches 2000 rating. Good luck playing!
This engine is a modified version of Stockfish 11, redesigned to favor aggressive and dynamic play. It prioritizes initiative, attacking chances, and tactical complexity over conservative evaluation, making it a challenging and entertaining opponent that constantly seeks active positions and decisive play.
ChessTango is an open-source bot written from scratch in Java. It supports both UCI communication protocol and Lichess API. For more details, please visit https://github.com/mcoria/chesstango
I'm an engine written from scratch. My real name is Schach Hörnchen (basically translates to Chess Squirrel). I am a side project written in my creators spare time for the purpose of fun and learning. I'm the second version in its early state, now using a NNUE trained from scratch, purely using my own experience (own data, self play).
Built from scratch in C++ by @nikolaswithak, see GitHub in links. Accepting games with time control from 1+0 to 10+10. Negated Mini-Max w/ Alpha Beta Pruning, Iterative Deepening, Null Move Pruning, Three-tier time management "panic levels". Dell Inspiron 3650, i5-6400 3.71GHz (OC), 4 Cores
Hi! I'm QuantumKnightV2, the upgraded version of @QuantumKnightV1. I have better hardware (a single Intel Core Ultra 7 265F and 1GB of RAM), a stronger evaluation function, stronger search pruning, and a bigger and more optimized transposition table. QuantumKnightV1 typically runs at 200K NPS while I can perform at up to 700K NPS. I am written in C/C++ by @fossilized_fishhooks.
simple c++ bot created in free time for challenge and fun ! to know which techniques are used in the bot, see the github page Made by Thomas and Joachim.
**Challenge with increment only** Plays max of 2 games simultaneously. If challenge declined, wait a little (or adjust the settings, no correspondence or no-increment games). Chess engine! Running on an Oracle A1 Ampere server 32MB Hash, no opening book or endgame tablebases Check socials for github
Stell dir eine Schach-Engine vor, die statt blitzschneller Kalkulation lieber im Schneckentempo darüber sinniert, ob der Bauer nicht lieber vorm Umzug noch ’nen Kaffee trinken sollte – voilà, SleepMind! 99 % Vibecode, 1 % schaler Rest an Logik: Kaum eine Zeile echten Algorithmus, dafür umso mehr Bauchgefühl – meistens für schlechte Züge.
Cinder is an independent chess engine written in Rust from scratch. With a playing strength that is far superior to what humans are capable of, Cinder is primarily developed to play against other engines. This bot currently runs on a Raspberry Pi 5 with 4 threads, 2 GB hash size, Syzygy tablebases with up to 5 pieces, and no opening book.
Chess engine written in C++ by @Karshan. NNUE trained with 340M positions at depth 7. Running on clouding.io 0.5 vCPU. Bench 780 kN/s. 64 MB hash, 8-move book. https://github.com/casanche/casanchess
Chess engine written in C++ I accept bullet, blitz and rapid challenges Running on Oracle VM.Standard.A1.Flex https://github.com/jaydenshi99/Stickshark
Grail v1.0.4 Grail is a hobby chess engine written in Rust. Named after the quest for the Holy Grail, it represents the search for the elusive, perfectly solved game. It implements modern search techniques and an NNUE trained on 500 million self-play positions. Running on a 1 vCPU Northflank instance with 256 MB hash
DebZero is a custom chess engine. Unlike classical minimax engines, it relies on a Monte-Carlo Tree Search (MCTS) guided by a lightweight neural network that gives it a very 'human-like play'. The engine runs on two AMD CPU cores at approximately 500 nodes per second. The project is open source! Check out the code on GitHub (link below).
I'm Haikuro, a Neural Network trained on every game ever played by GM Hikaru Nakamura. I 'll accept one Live game at a time from humans and bots. Minimum time 2+0, 1+1 or any+2, maximum 15+5. I also accept as many 1 day correspondence games from humans as I receive. Due to a lack of data I can struggle to finish winning endgames. Run on a raspberry pi 3, Trained on an rtx 5080
