ODU-Bench (Audio-Visual) - Key-Point Coverage: leaderboard

Metric: Key-point hit rate (%; share of reference key points on structured intent and required context covered by the predicted intent and context, demand scenes only; 1,347 audio-visual scenes (1,061 with a demand, 286 without; 1,070 synthetic and 277 human-recorded), user turns as video with audio, earlier assistant replies as text; key points judged by Qwen3.6-Flash with thinking). Source: arxiv.org. Saturation forecast: Around January 2028. 12 models tracked.

Top models

#ModelScore
1Qwen 3.5 Omni Plus65.8
2Seed 2.0 Lite64.1
3Gemini 3.1 Pro (Preview)63.4
4Gemini 3.7 Flash60.2
5Qwen3 Omni 30B A3B Instruct53.1
6Gemini 3.5 Flash Lite49.9
7Qwen2.5-Omni-7B42.1
8MiniCPM-o-4.535.5

Interactive version: theaggregate.ai/benchmark?slug=odu-bench-audio-visual-key-point-coverage · How It Works · Data refreshed daily, snapshot 2026-09-26.