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
Modelos de OpenAI por SI-IQ
44 modelos de OpenAI tienen resultados independientes de benchmarks en nuestros datos. El mejor clasificado es GPT-6 Astra, con un SI-IQ de 150.
Modelos
| 1 | GPT-6 Astra OpenAI | 150146–153 | 8 |
| 2 | GPT-6.1 Sol OpenAI | 148144–152 | 8 |
| 3 | GPT-6 Sol OpenAI | 140137–144 | 8 |
| 4 | GPT-5.6 Sol OpenAI | 140137–143 | 8 |
| 5 | GPT-5.5 OpenAI | 138135–141 | 9 |
| 6 | GPT-5.6 Terra OpenAI | 134131–137 | 8 |
| 7 | GPT-5.4 OpenAI | 131128–133 | 9 |
| 8 | GPT-5.6 Luna OpenAI | 130127–133 | 8 |
| 9 | GPT-6 Luna OpenAI | 130127–133 | 8 |
| 10 | GPT-5.2 OpenAI | 125122–128 | 10 |
| 11 | GPT-5 OpenAI | 121118–124 | 11 |
| 12 | GPT-5.1 OpenAI | 120115–125 | 7 |
| 13 | GPT-5.4 Nano OpenAI | 118115–121 | 8 |
| 14 | GPT-5.4 Mini OpenAI | 117113–120 | 8 |
| 15 | o3 OpenAI | 116113–119 | 10 |
| 16 | GPT-5 mini OpenAI | 115112–118 | 11 |
| 17 | o4-mini OpenAI | 112109–115 | 10 |
| 18 | gpt-oss-120babierto OpenAI | 10699–112 | 5 |
| 19 | GPT-5 nano OpenAI | 105102–109 | 10 |
| 20 | o3-mini OpenAI | 105101–108 | 9 |
| 21 | GPT-4.1 mini OpenAI | 9188–95 | 10 |
| 22 | GPT-4.1 OpenAI | 9187–94 | 10 |
| 23 | GPT-4.1 nano OpenAI | 7672–80 | 8 |
| 24 | GPT-4o (Aug 2024) OpenAI | 7368–77 | 8 |
| 25 | GPT-4o (Nov 2024) OpenAI | 6964–73 | 5 |
| 26 | GPT-4o mini OpenAI | 6864–73 | 9 |
| 27 | GPT-5.5 Pro OpenAI | 139135–142 | 5 datos insuficientes |
| 28 | GPT-5.4 Pro OpenAI | 135131–138 | 5 datos insuficientes |
| 29 | GPT-5.2 Pro OpenAI | 132128–135 | 3 datos insuficientes |
| 30 | GPT-5 Pro OpenAI | 124120–128 | 3 datos insuficientes |
| 31 | GPT-5.3 Codex OpenAI | 122108–136 | 2 datos insuficientes |
| 32 | o3-pro OpenAI | 119103–135 | 1 datos insuficientes |
| 33 | GPT-5.5 Instant OpenAI | 107103–111 | 4 datos insuficientes |
| 34 | o1 OpenAI | 10399–106 | 6 datos insuficientes |
| 35 | gpt-oss-20babierto OpenAI | 9589–101 | 3 datos insuficientes |
| 36 | o1-mini OpenAI | 9287–96 | 4 datos insuficientes |
| 37 | GPT-4.5 OpenAI | 8783–92 | 3 datos insuficientes |
| 38 | o1-preview OpenAI | 8480–89 | 3 datos insuficientes |
| 39 | GPT-4o (May 2024) OpenAI | 7066–75 | 4 datos insuficientes |
| 40 | GPT-4 Turbo (Apr 2024) OpenAI | 6864–73 | 4 datos insuficientes |
| 41 | GPT-4 (Jun 2023) OpenAI | 5751–63 | 4 datos insuficientes |
| 42 | GPT-4 (Mar 2023) OpenAI | 5240–65 | 2 datos insuficientes |
| 43 | GPT-3.5 Turbo (Nov 2023) OpenAI | 4943–55 | 2 datos insuficientes |
| 44 | GPT-3.5 Turbo (Jan 2024) OpenAI | 4742–53 | 5 datos insuficientes |
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); API list prices via LiteLLM's price table (MIT); each price links to the developer's pricing page; Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter..