SOCO - Text Query: leaderboard
SOCO (2026) tests whether models can match corresponding object parts across images, here posed to vision-language models as four-choice questions over candidate keypoints in a target image, scored strictly so that a question counts only if answered correctly under all four answer orders. In this board the query keypoint is given only as a text description, testing text-prompted part localization; accuracy in percent, higher is better.
Metric: Accuracy (%) when the query keypoint is given only as a text description (no source image), four-choice semantic-correspondence questions on SOCO image pairs (target image with candidate keypoints A to D), CircularEval: correct only if the model picks the right keypoint under all four answer permutations (random 0.4, consistent-permutation baseline 25); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 7 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 8B | 54 | #917 |
| 2 | Qwen 3 VL 4B | 44.4 | #1151 |
| 3 | InternVL3.5-8B | 39.6 | #1604 |
| 4 | Qwen 2.5 VL 7B Instruct | 39.1 | #1902 |
| 5 | GPT-4o | 37.6 | #751 |
| 6 | Qwen 2.5 VL 3B Instruct | 29.9 | #2713 |
| 7 | LLaVA-OneVision-7B | 24.3 | #2981 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=soco-text-query · How It Works · Data refreshed daily, snapshot 2026-11-21.