LMGame-Bench Tetris: leaderboard
GamingAgent places falling pieces and clears lines; native leaderboard game score. Primary GamingAgent harnessed cohort, distinct from model-only tab; full published score precision and source run annotations retained.
Metric: Score. Source: huggingface.co. Saturation forecast: Estimated already saturated. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4 | 125.67 | #181 |
| 2 | GPT-5 (High) | 84.33 | #98 (GPT-5) |
| 3 | O3 (2025-04-16) | 42 | #121 |
| 4 | O1 (2024-12-17) | 35 | #149 |
| 5 | DeepSeek R1 0528 | 33.67 | #222 |
| 6 | O4 Mini (2025-04-16) | 25.3 | #179 |
| 7 | Gemini 2.5 Pro (Preview 05-06) | 23.3 | #147 |
| 8 | Grok 3 Mini Beta | 21.3 | #296 |
| 9 | Claude Opus 4 (20250514) | 20 | #154 |
| 10 | GLM-4.5 | 19.67 | #275 |
| 11 | Claude Sonnet 4 (20250514) | 19.33 | #213 |
| 12 | Kimi K2 (0711) | 17 | #492 |
| 13 | Claude 3.7 Sonnet (20250219) | 16.3 | #201 |
| 14 | Gemini 2.5 Flash (Preview 04-17) | 16.3 | #237 |
| 15 | Claude 3.5 Sonnet (20241022) | 14.7 | #291 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=lmgame-bench-tetris · How It Works · Data refreshed daily, snapshot 2026-10-20.