benchmarks / metr-time-horizons
METR Time Horizons
time horizon (minutes). Status display_only. Contamination risk low. Maintainer METR. Source.
Display only.This benchmark cannot decide a crown under methodology v1.0.
| Model | Score | Date | Kind |
|---|---|---|---|
| Claude Opus 4.6 | 718.80683 | 2026-02-05 | observation |
| GPT-5.2 | 352.249302 | 2025-12-11 | observation |
| GPT-5.3 Codex | 349.530732 | 2026-02-05 | observation |
| Claude Opus 4.5 | 292.994594 | 2025-11-24 | observation |
| Gemini 3 Pro Preview | 224.325884 | 2025-11-18 | observation |
| GPT-5 | 203.012577 | 2025-08-07 | observation |
| O3 | 119.732634 | 2025-04-16 | observation |
| Claude Opus 4.1 | 100.472004 | 2025-08-05 | observation |
| Claude Opus 4 | 100.366123 | 2025-05-22 | observation |
| Claude 3.7 Sonnet | 60.388937 | 2025-02-24 | observation |
| GPT-4O | 6.991195 | 2024-05-13 | observation |
| GPT-4 0314 | 3.987428 | 2023-03-14 | observation |
| GPT-4 Turbo | 3.732787 | 2024-04-09 | observation |