HAVEN - Temporal Ordering (With Document): leaderboard
Temporal ordering on HAVEN (2026), a video understanding benchmark with aligned frames, shots, descriptions and full video documents, with the full video document added to the shuffled shot frames. The score is the Spearman rank correlation with the true order, from -1 to 1; higher is better.
Metric: Shot ordering: Spearman rank correlation (-1 to 1) between the predicted and the true order of shuffled video shots, frames plus the full video document, zero-shot on a random sample of 100 videos per source dataset; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 0.461 | #185 |
| 2 | GLM-4.6V-Flash | 0.306 | #1040 |
| 3 | Qwen 3 VL 8B Instruct | 0.3 | #1029 |
| 4 | InternVideo2.5-Chat-8B | 0.089 | #2036 |
| 5 | Qwen 2.5 VL 7B Instruct | 0.088 | #1902 |
| 6 | VideoLLaMA3-7B | 0.064 | #2050 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=haven-temporal-ordering-with-document · How It Works · Data refreshed daily, snapshot 2026-11-21.