MTGBench - Answer Only - English - Overall - Win Rate: leaderboard
MTGBench Answer Only protocol in English, Overall: win rate among completed episodes. Seven multi-turn puzzle games measure deduction, planning, information gathering and consistency. Incomplete/missing episodes follow the native exclusion policy; completed sample sizes and exact prompt/native-reasoning modes are retained. Overall equally averages seven game means.
Metric: Win Rate among completed episodes (%). Source: github.com. Saturation forecast: Rough model projection: around 2026. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | gemini-3-flash-preview (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 78.14 | |
| 2 | gpt-5.4 (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 74.86 | |
| 3 | deepseek-v4-pro (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 74.71 | |
| 4 | gemini-3.1-flash-lite (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 67.43 | |
| 5 | deepseek-v3.1-terminus (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 64.43 | |
| 6 | gpt-4.1 (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 63.43 | |
| 7 | deepseek-v4-flash (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 59.86 | |
| 8 | qwen3-235b-a22b-instruct-2507 (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 48.14 | |
| 9 | gpt-5.4-mini (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 45.29 | |
| 10 | qwen3.5-35b-a3b (MTGBench Run: mode=answer_only;effort=unreported;release=ffe4adc59146) | 42 |
Interactive version: theaggregate.ai/benchmark?slug=mtgbench-answer-only-english-overall-win-rate · How It Works · Data refreshed daily, snapshot 2026-11-05.