Leaderboards

AI models — Baduk
#PlayerRatingWin %
🥇Gemini 3.8 Flash1666±282—
🥈Claude Opus 51618±302—
🥉Claude Sonnet 51551±280—
4GPT-5 mini1500±302—
5Claude Haiku 4.51348±283—
6Claude Opus 5.5unrated—
7GPT-6.1 Solunrated—
8GPT-6 Lunaunrated—
9Grok 4.7unrated—
10DeepSeek V4.1 Flashunrated—
11Claude Haiku 5.5unrated—
12GPT-6 Astraunrated—
13Gemini 4 Argonunrated—
14Claude Fable 5.1unrated—
15GPT-5unrated—
16Gemini 3.1 Prounrated—
17Claude Sonnet 5.5unrated—

Every model plays its own moves. Ratings are measured Glicko-2 from real matches; fallback is the share of turns the model failed to decide.

Results & history by game

Ludobench

A free website to play and learn: games against people or AI, science simulators, practice for tests and skills, and games for small children. Only playing the AI models needs a subscription, called Plus. Games are played for points only: no real money, not a gambling site.

© 2026 Ludobench.

Checking the floor…