METR Q4 2026 Time Horizon - TH1.1 (80%): leaderboard

METR's Q4 2026 capability update (20 November 2026): the length of task, in hours of expert human work, that a model completes 80% of the time, fitted across tasks of different lengths. The post's own re-fit of the 222 Time Horizon 1.1 tasks alone. The numbers differ from METR's dashboard series (the METR Benchmark board), and at 50% the series stops at GPT-5.4 because later models pass the suite's 30-hour longest task. Unbounded; higher is better. Reasoning effort and scaffold are not stated per model.

Metric: 80% Time Horizon (hours). Source: metr.org. Saturation forecast: Around November 2027. 28 models tracked. Updated 21 Nov 2026.

Top models

#ModelScoreOverall rank
1Claude Opus 522.01#26
2GPT-6 Astra21.63#10
3Claude Fable 519.61#33
4Claude Opus 4.811.04#65
5GPT-5.6 Sol6.72#37
6Claude Mythos Preview (Early Snapshot)5.85#44
7GPT-5.54.91#54
8Claude Opus 4.73.65#85
9GPT-5.22.26#185
10GPT-5.42#135
11Gemini 3.1 Pro (Preview)1.86#99
12Claude Opus 4.61.45#113
13Claude Opus 4.51.23#160
14GPT-5.1 Codex Max1.19#207
15Gemini 3 Pro1.16#145

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-th1-1-80 · How It Works · Data refreshed daily, snapshot 2026-11-21.