Ranking de Tareas de agentes
Mira qué modelos de IA rinden mejor en Tareas de agentes, cuáles se mantienen fiables y dónde aparecen las mayores diferencias. Ordenar por: Costo total ↓.
Modelos mostrados
15
Promedio de Puntuación de Tareas de agentes
0.6
Mejor modelo
Claude Fable 5.2 1.0
514/514
Filtrar modelos
Ningún modelo coincide con la búsqueda y los filtros actuales.
| Rango | Modelo | Empresa | Puntuación de Tareas de agentes | Puntuación | Costo total | Pruebas correctas | Tiempo de respuesta (promedio) |
|---|---|---|---|---|---|---|---|
| #199 | DeepSeek V4 Pro 0423 none | DeepSeek | 1.0 | 0.7 | $0.083 | 1/1 | 65.5s |
| #366 | Nemotron 3 Super medium | NVIDIA | 0.5 | 0.6 | $0.081 | 0/1 | 138.3s |
| #335 | Gemini 3.1 Flash Lite minimal | 0.5 | 0.6 | $0.080 | 0/1 | 50.0s | |
| #478 | Command A+ high | Cohere | 0.4 | 0.3 | $0.079 | 0/1 | 156.0s |
| #304 | Solar Pro 4 high | Upstage | 0.3 | 0.6 | $0.077 | 0/1 | 312.3s |
| #392 | KAT Coder AIR V2.5 high | Kwaipilot | 0.3 | 0.5 | $0.077 | 0/1 | 1.10s |
| #266 | Qwen3.7 Flash medium | Qwen | 0.5 | 0.7 | $0.076 | 0/1 | 149.0s |
| #176 | GPT-6 Luna high | OpenAI | 0.5 | 0.8 | $0.074 | 0/1 | 82.8s |
| #482 | Command A+ low | Cohere | 0.3 | 0.3 | $0.073 | 0/1 | 128.5s |
| #212 | Qwen3.8 27B medium | Qwen | 0.5 | 0.7 | ~$0.073 | 0/1 | 187.9s |
| #181 | Qwen3.7 Flash low | Qwen | 1.0 | 0.8 | $0.071 | 1/1 | 106.7s |
| #424 | Qwen3 Coder Next medium | Qwen | 0.5 | 0.5 | $0.071 | 0/1 | 151.4s |
| #439 | KAT Coder AIR V2.5 default | Kwaipilot | 0.3 | 0.4 | $0.070 | 0/1 | 1.07s |
| #337 | Gemini 3.1 Flash Lite none | 0.5 | 0.6 | $0.069 | 0/1 | 43.8s | |
| #455 | Grok 4.1 Fast medium | X AI | 0.0 | 0.4 | $0.069 | 0/0 | 0ms |