IPIBench - Proactive Monitoring: leaderboard

IPIBench (2026) streams videos to multimodal LLMs at one frame per second and checks whether they respond proactively at the right moment, manage monitoring tasks the user cancels or changes, and handle reactive questions interleaved with proactive tasks. This board is accuracy on the proactive monitoring tasks (mean of timing, understanding and repeated proactiveness); percent, higher is better.

Metric: Accuracy (%) on the proactive monitoring tasks (mean of timing, understanding and repeated proactiveness), on IPIBench (1,831 videos, 3,738 interaction instances from egocentric and exocentric datasets) under a unified streaming protocol at 1 FPS (offline models see a sliding window of the latest 16 frames); a proactive response counts only if the first trigger falls within one second of the ground-truth moment, multi-turn tasks only if every required trigger and answer is correct, reactive answers by bidirectional substring matching against up to 20 reference answers; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 16 models tracked. Updated 14 Nov 2026.

Top models

#ModelScoreOverall rank
1GPT-4o30.09#751
2GLM-4.6V29.62#710
3GPT-5.429.04#135
4Gemini 3 Pro28.93#145
5Gemini 2.5 Pro27.94#276
6Qwen 3.5 9B (Harness: AURA2 IPIBench offline30frames; Thinking: On)26.65
7Qwen 3 VL 8B23.96#917
8Qwen 3.5 Plus22.5#242
9InternVL3-8B19.53#1782
10MOSS-VL-Realtime11.51
11Dispider7.06#3815
12LLaVA-OneVision-7B6.93#2981
13VideoLLM-online-8B5.53
14Qwen 3.5 9B (Harness: AURA2 IPIBench offline30frames; Thinking: Off)3.82
15JoyAI-VL-Interaction3.41

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=ipibench-proactive-monitoring · How It Works · Data refreshed daily, snapshot 2026-11-21.