Skip to content
BoardBench
Games
Overall
Yatzi
Scotland Yard
Splendor
Models
Anthropic
Anthropic
Claude Fable 5 Medium
Claude Opus 5 Medium
Claude Sonnet 5 Medium
OpenAI
OpenAI
GPT-5.6 Luna High
GPT-5.6 Sol Medium
xAI
xAI
Grok 4.5 High
Grok 4.5 Medium
Grok 4.5 Low
Grok 4.6 High
Grok 4.6 Medium
DeepSeek
DeepSeek
DeepSeek V4 Flash 0731 High
Moonshot
Moonshot
Kimi K3 Medium
Methodology
Blog
Appearance
System
Light
Dark
Blog
16 Aug 2026
Introducing BoardBench
A benchmark to determine an AI's capability to play board games
Look
Slate
Felt
Card
Night
Linen
Photo
Paper
Shards
Hide