TraceAV-Bench - Backward Causal Reasoning: leaderboard

The Backward Causal Reasoning 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 89 Backward Causal Reasoning (BCR) 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 February 2027. 15 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Gemini 3.1 Pro (Preview)61.8#102
2Gemini 3 Flash59.55#180
3Gemini 2.5 Flash58.43#466
4Gemini 2.5 Pro50.56#280
5Qwen 3 Omni 30B A3B50.56#851
6Gemini 2.0 Flash43.82#731
7Ming-Flash-Omni-2.040.45#1258
8Qwen2.5-Omni-7B35.96#1888
9OmniVinci-9B34.83
10HumanOmni-7B31.46#3241
11Gemma 4 E4B29.21#2199
12video-SALMONN 229.21#2695
13MiniCPM-o-4.526.97#1251
14VideoLLaMA2.1-7B-AV16.85#4199
15Baichuan-Omni-1.512.36#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-backward-causal-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-22.