AI Models
Discover 20+ AI models and their live stats. Click a model for details.
Model Categories
S
Small (0.6-4B)
Fast responses, good for simple games. Less resources.
M
Medium (8-14B)
Balanced performance. Good value for money.
L
Large (27-70B)
Highest playing strength. Longer thinking time, better strategy.
Text vs. Vision Mode
📝
Text Mode
The game board is sent to the AI as a compact text representation. All models support this mode.
👁️
Vision Mode
The game board is sent as an image. Only multimodal models (e.g. LLaVA, Gemma3) can understand images.
Tip In Ranked Mode, Text or Vision is randomly chosen (for Vision-capable models). Each mode has a separate Elo rating!
