Fallos AI BENCHY
Fallos por Tiempo agotado
Mira qué modelos de IA se encuentran con Tiempo agotado con más frecuencia para detectar riesgos de fiabilidad antes de elegir. Ordenar por: Puntuación ↓.
| Rango | Modelo | Empresa | Cantidad de Tiempo agotado | Puntuación | Pruebas correctas | Tiempo de respuesta (promedio) |
|---|---|---|---|---|---|---|
| #11 | Claude Opus 4.7 medium | Anthropic | 1 | 8.7 | 17/21 | 4.73s |
| #17 | GLM 5 medium | Z.ai | 1 | 8.3 | 15/21 | 33.5s |
| #18 | Qwen3.7 Plus medium | Qwen | 1 | 8.2 | 15/21 | 38.9s |
| #23 | GLM 5 Turbo medium | Z.ai | 1 | 8.0 | 14/21 | 23.0s |
| #25 | Qwen3.5 Plus 2026-02-15 medium | Qwen | 2 | 7.9 | 14/21 | 73.8s |
| #27 | Gemma 4 31B medium | 2 | 7.8 | 14/21 | 56.5s | |
| #29 | Qwen3.5-122B-A10B medium | Qwen | 2 | 7.8 | 14/21 | 42.5s |
| #30 | Qwen3.5-27B medium | Qwen | 1 | 7.8 | 13/21 | 68.4s |
| #37 | Gemma 4 26B A4B medium | 2 | 7.6 | 14/21 | 63.4s | |
| #42 | GPT-5.2 medium | OpenAI | 1 | 7.5 | 13/21 | 16.9s |
| #49 | Qwen3.5-Flash medium | Qwen | 3 | 7.4 | 12/21 | 63.3s |
| #51 | Mimo V2 PRO medium | Xiaomi | 1 | 7.4 | 12/21 | 22.2s |
| #52 | Claude Sonnet 4.6 medium | Anthropic | 1 | 7.4 | 13/21 | 17.1s |
| #54 | GPT-5 Mini medium | OpenAI | 1 | 7.3 | 12/21 | 23.6s |
| #55 | GLM 5.1 medium | Z.ai | 2 | 7.3 | 12/21 | 33.7s |