Página automatizada. Redactada y traducida automáticamente por IA y verificada con las fuentes mediante software; sin editor humano. SI-IQ es un índice elaborado a partir de resultados públicos de benchmarks, no una prueba de CI. Cómo funciona
Mystery Game Puzzles: resultados de modelos de IA
Mystery Game Puzzles es un benchmark de Razonamiento ejecutado por Epoch AI. En nuestros datos hay resultados publicados de 78 modelos.
Forma parte del índice SI-IQ (carga α = 1,15, nivel de azar 9 %).
- Evaluador
- Epoch AI
- Página principal del benchmark
- epoch.ai/benchmarks/mystery-game-puzzles
Resultados
| 1 | GPT-6 Astra OpenAI | 84,0 % | 150146–153 | 9 |
| 2 | GPT-6.1 Sol OpenAI | 80,0 % | 148144–152 | 9 |
| 3 | Claude Opus 5.5 Anthropic | 71,0 % | 144141–148 | 9 |
| 4 | Claude Sonnet 5.5 Anthropic | 65,0 % | 140137–143 | 9 |
| 5 | Gemini 4 Argon Google DeepMind | 64,0 % | 140137–143 | 9 |
| 6 | Claude Opus 5 Anthropic | 59,0 % | 136133–140 | 9 |
| 7 | Claude Fable 5.1 Anthropic | 58,0 % | 140136–143 | 8 |
| 8 | GPT-5.6 Sol OpenAI | 58,0 % | 140137–143 | 9 |
| 9 | GPT-5.5 OpenAI | 56,0 % | 138135–141 | 10 |
| 10 | GPT-6 Sol OpenAI | 56,0 % | 140137–144 | 9 |
| 11 | Claude Fable 5 Anthropic | 52,0 % | 138134–141 | 9 |
| 12 | Muse Wonder Meta AI | 51,0 % | 136133–139 | 4 resultados de menos de 5 benchmarks; resultados en menos de 3 áreas; resultados de menos de 2 evaluadores |
| 13 | Gemini 3.8 Flash Google DeepMind | 47,0 % | 130127–133 | 9 |
| 14 | DeepSeek V4 Pro 0813abierto DeepSeek | 43,0 % | 128125–131 | 9 |
| 15 | DeepSeek V4.1 Flashabierto DeepSeek | 43,0 % | 128125–131 | 9 |
| 16 | Qwen 3.8 Max Alibaba | 38,0 % | 130127–133 | 7 resultados de menos de 2 evaluadores |
| 17 | GPT-5.4 OpenAI | 37,0 % | 131128–133 | 10 |
| 18 | Gemini 3.7 Flash Google DeepMind | 37,0 % | 131128–134 | 9 |
| 19 | Claude Opus 4.8 Anthropic | 36,0 % | 132129–135 | 9 |
| 20 | Claude Sonnet 5 Anthropic | 35,0 % | 126123–129 | 9 |
| 21 | GPT-5.6 Terra OpenAI | 35,0 % | 134131–137 | 9 |
| 22 | DeepSeek V4 Flash 0731abierto DeepSeek | 34,0 % | 124121–127 | 9 |
| 23 | Gemini 3.1 Pro Google DeepMind | 34,0 % | 127124–130 | 10 |
| 24 | Grok 4.6 xAI | 34,0 % | 126123–129 | 9 |
| 25 | GLM-5.3abierto Z.ai (Zhipu AI) | 33,0 % | 125122–128 | 9 |
| 26 | Gemini 3.5 Flash Google DeepMind | 32,0 % | 126123–129 | 10 |
| 27 | Qwen3.7-Max Alibaba | 32,0 % | 125122–128 | 10 |
| 28 | Claude Haiku 5.5 Anthropic | 30,0 % | 129126–132 | 9 |
| 29 | Gemini 3.6 Flash Google DeepMind | 30,0 % | 125122–128 | 9 |
| 30 | Grok 4.7 xAI | 29,0 % | 123120–127 | 7 resultados de menos de 2 evaluadores |
| 31 | o3 OpenAI | 29,0 % | 116113–119 | 10 |
| 32 | Claude Opus 4.7 Anthropic | 28,0 % | 127124–130 | 10 |
| 33 | Gemini 3 Flash Google DeepMind | 26,0 % | 122119–125 | 10 |
| 34 | Kimi K3abierto Moonshot | 26,0 % | 127124–130 | 9 |
| 35 | Claude Opus 4.6 Anthropic | 25,0 % | 123120–126 | 10 |
| 36 | Muse Spark 1.3 Meta AI | 25,0 % | 128125–131 | 5 resultados en menos de 3 áreas; resultados de menos de 2 evaluadores |
| 37 | GPT-5 OpenAI | 23,0 % | 121118–124 | 12 |
| 38 | GPT-5.2 OpenAI | 23,0 % | 125122–128 | 11 |
| 39 | Claude Opus 4.5 Anthropic | 22,0 % | 115111–118 | 11 |
| 40 | Qwen 3.6 35B-A3Babierto Alibaba | 22,0 % | 111107–115 | 6 |
| 41 | Claude Opus 4.1 Anthropic | 21,0 % | 105102–109 | 6 resultados de menos de 2 evaluadores |
| 42 | GPT-5.6 Luna OpenAI | 21,0 % | 130127–133 | 9 |
| 43 | Nemotron 3 Ultraabierto Nvidia | 20,0 % | 114108–120 | 4 resultados de menos de 5 benchmarks |
| 44 | Qwen 3.5 Flash (hosted 35B-A3B) Alibaba | 20,0 % | 107103–111 | 6 resultados de menos de 2 evaluadores |
| 45 | GLM-5.2abierto Z.ai (Zhipu AI) | 19,0 % | 121118–124 | 10 |
| 46 | GPT-5.1 OpenAI | 19,0 % | 120115–125 | 8 |
| 47 | Gemini 3.5 Flash-Lite Google DeepMind | 19,0 % | 109106–113 | 8 |
| 48 | Qwen 3.6 Max (Preview) Alibaba | 19,0 % | 118112–124 | 5 resultados de menos de 2 evaluadores |
| 49 | Kimi K2.6abierto Moonshot | 18,0 % | 121118–124 | 10 |
| 50 | Qwen 3.6 Flash Alibaba | 18,0 % | 107103–111 | 6 resultados de menos de 2 evaluadores |
| 51 | Qwen3.5 397B-A17Babierto Alibaba | 18,0 % | 114109–118 | 6 |
| 52 | Claude Sonnet 4.5 Anthropic | 17,0 % | 110107–113 | 12 |
| 53 | DeepSeek-V4-Proabierto DeepSeek | 17,0 % | 117113–120 | 10 |
| 54 | Mistral Large 4 Mistral AI | 17,0 % | 122119–125 | 7 resultados de menos de 2 evaluadores |
| 55 | Qwen 3.5 Plus (hosted 397B-A17B) Alibaba | 17,0 % | 109103–115 | 5 resultados de menos de 2 evaluadores |
| 56 | Qwen3.5-122B-A10Babierto Alibaba | 17,0 % | 112103–121 | 3 resultados de menos de 5 benchmarks |
| 57 | Qwen3.7-Plus Alibaba | 17,0 % | 115110–119 | 6 |
| 58 | Claude Sonnet 4.6 Anthropic | 16,0 % | 113108–118 | 8 |
| 59 | Qwen3.7 Flash Alibaba | 15,0 % | 107103–112 | 4 resultados de menos de 5 benchmarks; resultados de menos de 2 evaluadores |
| 60 | GPT-4 (Jun 2023) OpenAI | 12,0 % | 5751–63 | 4 resultados de menos de 5 benchmarks; resultados de menos de 2 evaluadores |
| 61 | GPT-4o mini OpenAI | 12,0 % | 6864–73 | 10 |
| 62 | Qwen 3.6 Plus Alibaba | 12,0 % | 113109–117 | 7 resultados de menos de 2 evaluadores |
| 63 | GPT-5.4 Mini OpenAI | 11,0 % | 117113–120 | 9 |
| 64 | GPT-5 mini OpenAI | 10,0 % | 115112–118 | 12 |
| 65 | GPT-5 nano OpenAI | 9,0 % | 105102–109 | 11 |
| 66 | GPT-5.4 Nano OpenAI | 9,0 % | 118115–121 | 9 |
| 67 | Qwen3-235B-A22B-Thinking (Jul 2025)abierto Alibaba | 9,0 % | 111105–118 | 6 |
| 68 | GLM-5.3-Flashabierto Z.ai (Zhipu AI) | 8,0 % | 120117–124 | 8 |
| 69 | MiniMax-M3abierto MiniMax | 8,0 % | 108102–114 | 6 |
| 70 | GPT-4.1 mini OpenAI | 7,0 % | 9188–95 | 11 |
| 71 | GPT-6 Luna OpenAI | 7,0 % | 130127–133 | 9 |
| 72 | Qwen3.6 27Babierto Alibaba | 7,0 % | 115110–119 | 6 |
| 73 | o3-mini OpenAI | 7,0 % | 105101–108 | 9 |
| 74 | Inkling-Smallabierto Thinking Machines | 6,0 % | 118115–122 | 8 |
| 75 | Qwen3-Max Alibaba | 5,0 % | 106103–110 | 6 resultados de menos de 2 evaluadores |
| 76 | o4-mini OpenAI | 5,0 % | 112109–115 | 10 |
| 77 | GPT-3.5 Turbo (Jan 2024) OpenAI | 3,0 % | 4742–53 | 5 resultados de menos de 2 evaluadores |
| 78 | gpt-oss-120babierto OpenAI | 2,0 % | 10699–112 | 6 |
Fuentes de datos: Aider polyglot leaderboard data (Apache-2.0); Arena leaderboard dataset on Hugging Face (CC BY 4.0); Berkeley Function Calling Leaderboard, Gorilla project (Apache-2.0); Epoch AI, ‘Capabilities & benchmarking’ (CC BY 4.0); LiteLLM, model_prices_and_context_window.json (MIT); Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter (licencias y cambios).