LLM Game Benchmark: leaderboard
Grid-based game competition benchmark evaluating LLM strategic play and invalid-move behavior in Tic-Tac-Toe, Connect Four, and Gomoku.
Metric: Source Win Rate (imported). Source: research-outcome.github.io. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 42.17 | #299 |
| 2 | GPT-4o | 40.33 | #343 |
| 3 | Gemini 1.5 Pro | 35.17 | #445 |
| 4 | GPT-4 Turbo | 35 | #404 |
| 5 | Gemini 1.5 Flash | 32.5 | #618 |
| 6 | Claude 3 Sonnet (20240229) | 32.33 | #677 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-game-benchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.