HoloCount - Complementary Exclusion: leaderboard
Counting the objects that belong to one set but not another: a subset of HoloCount (Meituan, 2026), a visual counting benchmark for multimodal LLMs. Models answer with a single integer, zero-shot; exact-match accuracy in percent, higher is better.
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 objects in one set but not another (101 questions); higher is better). Source: arxiv.org. Saturation forecast: Around February 2027. 30 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 84.2 | #94 |
| 2 | Qwen 3.5 397B A17B | 80.2 | #246 |
| 3 | Gemini 3 Flash (Preview) | 80.2 | #144 |
| 4 | Qwen 3.5 122B A10B | 79.2 | #331 |
| 5 | Qwen 3.5 35B A3B | 78.2 | #482 |
| 6 | Qwen 3.5 27B | 76.2 | #359 |
| 7 | Gemini 2.5 Pro | 74.3 | #266 |
| 8 | Kimi K2.6 | 74.3 | #178 |
| 9 | Qwen 3.5 4B | 70.3 | #1175 |
| 10 | Qwen 3.5 9B | 67.3 | #803 |
| 11 | GPT-5.5 | 66.3 | #49 |
| 12 | Kimi K2.5 | 63.4 | #256 |
| 13 | Claude Opus 4.8 | 63.4 | #56 |
| 14 | GPT-5.4 | 61.4 | #123 |
| 15 | GPT-4.1 | 60.4 | #495 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=holocount-complementary-exclusion · How It Works · Data refreshed daily, snapshot 2026-11-13.