MiGUE-Bench - Temporal Ordering (End): leaderboard
The temporal ordering task of MiGUE-Bench (2026), a benchmark of event analysis over news documents: accuracy in percent at ordering the events of 150 multi-document event sets by their end, where the model outputs a full permutation. Higher is better.
Metric: Accuracy (%) of the event-end ordering over 150 multi-document event sets (output a full permutation); higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 Pro | 83.33 | #124 |
| 2 | Claude Opus 4.5 | 80.79 | #156 |
| 3 | DeepSeek V3.2 | 63 | #406 |
| 4 | GLM-4.7 | 56 | #366 |
| 5 | Gemini 3 Pro | 37.75 | #137 |
| 6 | Kimi K2 | 19.87 | #439 |
| 7 | Qwen 3 Max | 19.87 | #403 |
| 8 | Claude Haiku 4.5 | 17.22 | #568 |
| 9 | Qwen 3 235B A22B | 12.58 | #623 |
| 10 | Qwen 3 8B | 7.95 | #1830 |
| 11 | Qwen 3 30B A3B | 5.3 | #1122 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=migue-bench-temporal-ordering-end · How It Works · Data refreshed daily, snapshot 2026-11-13.