SYNCR - Object Counting: leaderboard
Accuracy (%) on counting distinct objects seen across several videos in SYNCR (2026), a synthetic benchmark of multi-video questions with programmatically verified answers; zero-shot four-option multiple choice, exact match, 200 questions per task (chance 25). Higher is better.
Metric: Exact-match accuracy (%) on the SYNCR Object Counting task (counting distinct objects seen across several videos), zero-shot four-option multiple choice over 2 to 4 labelled input videos, exact match, deterministic decoding, on the paper's balanced subset of 200 questions per task; chance is 25; higher is better. Source: arxiv.org. Saturation forecast: Around January 2028. 18 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 68 | #135 |
| 2 | InternVL3.5-38B | 66.5 | #1021 |
| 3 | LLaVA-OV-72B | 65 | #1879 |
| 4 | Qwen 3 VL 8B Instruct | 60.5 | #1029 |
| 5 | InternVL3.5-14B | 58.5 | #1504 |
| 6 | Qwen 3 VL 4B Instruct | 56 | #1332 |
| 7 | Qwen 3 VL 32B Instruct | 54.5 | #606 |
| 8 | InternVL3.5-8B | 51.5 | #1604 |
| 9 | InternVL3.5-4B | 51.5 | #1930 |
| 10 | InternVL3.5-2B | 50.5 | #3284 |
| 11 | Qwen 2.5 VL 32B Instruct | 50 | #1062 |
| 12 | Qwen 2.5 VL 7B Instruct | 48.5 | #1902 |
| 13 | Gemini 3 Flash | 48 | #175 |
| 14 | Qwen 2.5 VL 72B Instruct | 46.5 | #841 |
| 15 | Qwen 3 VL 2B Instruct | 46.5 | #2367 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=syncr-object-counting · How It Works · Data refreshed daily, snapshot 2026-11-21.