Mejor IA para programación
Las clasificaciones combinan varias puntuaciones de benchmarks ponderadas según su relevancia para cada caso de uso. Actualizado agosto 2026.
Respuesta rápida
1. Claude Fable 5.1 (Anthropic, Puntuación: 94.1), 2. Claude Fable 5 (Anthropic, Puntuación: 85.6), 3. Claude Fable 5.1 (XHigh) (Anthropic, Puntuación: 85.5).
| # | Modelo | Proveedor | Puntuación | Velocidad | Precio (salida) |
|---|---|---|---|---|---|
| 1 | Claude Fable 5.1 | Anthropic | 94.1 | 85 t/s | $15/M |
| 2 | Claude Fable 5 | Anthropic | 85.6 | 71 t/s | $50/M |
| 3 | Claude Fable 5.1 (XHigh) | Anthropic | 85.5 | 78 t/s | $15/M |
| 4 | Claude Fable 5.1 (High) | Anthropic | 85.5 | 92 t/s | $15/M |
| 5 | Claude Fable 5.1 (Medium) | Anthropic | 85.5 | 104 t/s | $15/M |
| 6 | Claude Fable 5.1 (Low) | Anthropic | 85.5 | 118 t/s | $15/M |
| 7 | Claude Opus 5.5 | Anthropic | 84.3 | 68 t/s | $25/M |
| 8 | GPT-6 Astra | OpenAI | 76.9 | 82 t/s | $50/M |
| 9 | Kimi K3 | Moonshot AI | 76.7 | 37 t/s | $15/M |
| 10 | Claude Opus 5.5 (XHigh) | Anthropic | 74.9 | 62 t/s | $25/M |
| 11 | Claude Opus 5.5 (High) | Anthropic | 74.9 | 74 t/s | $25/M |
| 12 | Claude Opus 5.5 (Medium) | Anthropic | 74.9 | 82 t/s | $25/M |
| 13 | Claude Opus 5.5 (Low) | Anthropic | 74.9 | 92 t/s | $25/M |
| 14 | GPT-6 Astra (xhigh) | OpenAI | 70.6 | 76 t/s | $50/M |
| 15 | GPT-6 Astra (high) | OpenAI | 70.6 | 90 t/s | $50/M |
Claude Fable 5.1
GPT-6 Astra
Kimi K3