METR Q4 2026 Time Horizon - TH1.1 + TH2 (50%): leaderboard
METR's Q4 2026 capability update (20 November 2026): the length of task, in hours of expert human work, that a model completes 50% of the time, fitted across tasks of different lengths. This is the post's headline series: the 222 older Time Horizon 1.1 tasks pooled with TH2, METR's new suite of 36 longer tasks (up to 200 human hours) scored against several human (score, time) points per task, for the 15 models run on both. Unbounded; higher is better. Reasoning effort and scaffold are not stated per model.
Metric: 50% Time Horizon (hours). Source: metr.org. Saturation forecast: Around December 2027. 15 models tracked. Updated 21 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 40.27 | #10 |
| 2 | Claude Opus 5.5 | 38.96 | #9 |
| 3 | Claude Fable 5.1 | 30.21 | #20 |
| 4 | Claude Opus 5 | 19.91 | #26 |
| 5 | Claude Fable 5 | 15.58 | #33 |
| 6 | GPT-5.6 Sol | 11.94 | #37 |
| 7 | Claude Opus 4.8 | 11.84 | #65 |
| 8 | GPT-5.5 | 11.07 | #54 |
| 9 | Claude Opus 4.7 | 8.38 | #85 |
| 10 | Claude Opus 4.6 | 6.4 | #113 |
| 11 | GPT-5.4 | 5.94 | #135 |
| 12 | Claude Opus 4.5 | 4.59 | #160 |
| 13 | GPT-5 | 3.01 | #178 |
| 14 | O1 | 0.5875 | #403 |
| 15 | GPT-4o | 0.1185 | #751 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=metr-q4-2026-time-horizon-th1-1-plus-th2-50 · How It Works · Data refreshed daily, snapshot 2026-11-21.