TraceAV-Bench - Temporal Sequencing: leaderboard

The Temporal Sequencing sub-task of TraceAV-Bench (2026), part of its audio-visual joint reasoning dimension: multiple-choice questions over long videos whose answers chain several timestamped evidence hops. The score is accuracy in percent with full audio and video input; higher is better.

Metric: Accuracy (%) on the 97 Temporal Sequencing (TS) questions, four-option multiple choice over long audio-visual videos (10 to 140 minutes), exact match of the selected option set, each model at its official inference setting, with the full audio and video input; higher is better. Source: arxiv.org. Saturation forecast: Around March 2027. 15 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro63.92#280
2Gemini 3.1 Pro (Preview)60.82#102
3Gemini 2.5 Flash58.76#466
4Gemini 3 Flash53.61#180
5Gemini 2.0 Flash53.61#731
6Ming-Flash-Omni-2.053.61#1258
7Qwen 3 Omni 30B A3B51.55#851
8OmniVinci-9B44.33
9video-SALMONN 241.24#2695
10Gemma 4 E4B38.14#2199
11MiniCPM-o-4.536.08#1251
12Qwen2.5-Omni-7B32.99#1888
13HumanOmni-7B31.96#3241
14VideoLLaMA2.1-7B-AV29.9#4199
15Baichuan-Omni-1.514.43#3189

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=traceav-bench-temporal-sequencing · How It Works · Data refreshed daily, snapshot 2026-11-22.