ContrAR (Multiple Frames) - Outdoor Navigation: leaderboard

ContrAR (2026) tests whether vision-language models detect contradictory virtual content attacks in augmented reality, answering yes or no on real AR videos of which half carry an overlay that contradicts the physical scene, given the first, middle and last frames of each video. Detection accuracy in percent on the outdoor navigation videos; higher is better.

Metric: Detection accuracy (%) for contradictory virtual content attacks in AR videos (60 outdoor navigation videos; half the videos carry a contradiction, a balanced yes/no decision), first, middle and last frames given (multi-frame prompting); mean of three runs; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 11 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-4o90#751
2GPT-588.33#178
3GPT-4.185#505
4Gemini 2.5 Pro81.67#276
5Gemini 2.5 Flash76.67#463
6Qwen 2.5 VL 72B Instruct76.67#841
7Grok 2 Vision73.33#1437
8Claude Sonnet 4.571.67#261
9Grok 470#319
10Qwen 2.5 VL 7B Instruct66.67#1902
11Claude Haiku 4.558.33#583

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

Interactive version: theaggregate.ai/benchmark?slug=contrar-multiple-frames-outdoor-navigation · How It Works · Data refreshed daily, snapshot 2026-11-21.