HoloCount - Partial Object Occlusion: leaderboard

Metric: Accuracy (%; exact-match accuracy of the single integer count a multimodal model returns zero-shot for an image and a counting question, with a strict integer-only system prompt; counting heavily occluded objects (103 questions); higher is better). Source: arxiv.org. Saturation forecast: Around 2029. 30 models tracked.

Top models

#ModelScore
1Qwen 3.5 35B A3B65
2Qwen 3.5 9B64.1
3Qwen 3.5 122B A10B64.1
4Gemini 3.1 Pro (Preview)63.1
5Kimi K2.661.2
6Qwen 3.5 397B A17B61.2
7Qwen 3.5 27B61.2
8Gemini 3 Flash (Preview)61.2
9Qwen 2.5 VL 72B Instruct58.3
10Gemini 2.5 Pro56.3
11Claude Opus 4.855.3
12Qwen 3.5 4B55.3
13GPT-4.150.5
14Claude Opus 4.750.5
15Qwen 3 VL 32B Instruct47.6

Interactive version: theaggregate.ai/benchmark?slug=holocount-partial-object-occlusion · How It Works · Data refreshed daily, snapshot 2026-09-29.