MapVerse - Counting: leaderboard
Metric: Exact match (%) on MapVerse's 2,395 human-authored counting questions over real-world maps, zero-shot image plus question; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 VL 7B Instruct | 23 | #643 |
| 2 | InternVL3-8B | 21.1 | #606 |
| 3 | Mistral Small 3.2 | 19.8 | #587 |
| 4 | Llama 3.2 11B Instruct | 19.2 | #1112 |
| 5 | aya-vision-8B | 16.8 | #1094 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=mapverse-counting · How It Works · Data refreshed daily, snapshot 2026-10-11.