GPT-4.1 vs gpt-oss-120b
Lado a lado. Cada métrica. Cada benchmark.
| Tipo | GPT-4.1 | gpt-oss-120b |
|---|---|---|
| Provider |  OpenAI | OpenAI |
| puntuación promedio | 43.3 | 46.9 |
| Precio de entrada | $2.00 | $0.04 |
| Precio de salida | $8.00 | $0.19 |
| Ventana de contexto | 1.0M tokens (~524 books) | 131K tokens (~66 books) |
| Publicado el | 2025-04-14 | 2025-08-05 |
| Código abierto | Proprietary | Open Source |
Puntuaciones de benchmark
12 benchmarks · GPT-4.1: 7, gpt-oss-120b: 5
| Benchmark | Categoría | GPT-4.1 | gpt-oss-120b |
|---|---|---|---|
| Aider polyglot | coding | 52.4 | 41.8 |
| Fiction.LiveBench | knowledge | 63.9 | 44.4 |
| GPQA diamond | knowledge | 55.9 | 67.7 |
| HELM — GPQA | knowledge | 65.9 | 68.4 |
| HELM — IFEval | language | 83.8 | 83.6 |
| HELM — MMLU-Pro | knowledge | 81.1 | 79.5 |
| HELM — Omni-MATH | math | 47.1 | 68.8 |
| HELM — WildBench | reasoning | 85.4 | 84.5 |
| OTIS Mock AIME 2024-2025 | math | 38.3 | 88.9 |
| SimpleBench | reasoning | 12.4 | 6.5 |
| SWE-Bench Verified (Bash Only) | coding | 39.6 | 26.0 |
| WeirdML | coding | 39.0 | 48.2 |