GPT-5.4 Pro vs Claude Opus 4.6
Lado a lado. Cada métrica. Cada benchmark.
| Tipo | GPT-5.4 Pro | Claude Opus 4.6 |
|---|---|---|
| Provider |  OpenAI |  Anthropic |
| puntuación promedio | 66.7 | 57.5 |
| Precio de entrada | $30.00 | $5.00 |
| Precio de salida | $180.00 | $25.00 |
| Ventana de contexto | 1.1M tokens (~525 books) | 1.0M tokens (~500 books) |
| Publicado el | 2026-03-05 | 2026-02-04 |
| Código abierto | Proprietary | Proprietary |
Puntuaciones de benchmark
8 benchmarks · GPT-5.4 Pro: 8, Claude Opus 4.6: 0
| Benchmark | Categoría | GPT-5.4 Pro | Claude Opus 4.6 |
|---|---|---|---|
| ARC-AGI | reasoning | 94.5 | 94.0 |
| ARC-AGI-2 | reasoning | 83.3 | 69.2 |
| Chess Puzzles | knowledge | 58.6 | 17.0 |
| FrontierMath-2025-02-28-Private | math | 50.0 | 40.7 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 37.5 | 22.9 |
| GPQA diamond | knowledge | 92.8 | 87.4 |
| SimpleBench | reasoning | 68.9 | 61.1 |
| SimpleQA Verified | knowledge | 47.8 | 46.5 |