TsuGO: leaderboard
First-move accuracy on the Easy tier with four supplied candidate moves (K=4), averaged across available input modalities. Open-search and other difficulty tiers are separate settings.
Metric: First-move Accuracy (%, Easy, K=4, modality average). Source: arxiv.org. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 80 | #58 |
| 2 | Kimi K2.5 | 52 | #144 |
| 3 | DeepSeek V3.2 | 39.3 | #205 |
| 4 | MiniMax-M2.5 | 37.7 | #303 |
| 5 | DeepSeek R1 0528 | 32.3 | #222 |
| 6 | Gemini 2.5 Flash | 32 | #245 |
| 7 | GLM-4.6V | 31 | #316 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=tsugo · How It Works · Data refreshed daily, snapshot 2026-10-20.