CoreCognition Single Image Toolusing: leaderboard

CoreCognition Single Image Toolusing — CoreCognition native component accuracy and reported mean. Image+Video, image-only and single-image populations are separate; no pooling or replacement across cohorts. Metric: Accuracy (%). Source: https://grow-ai-like-a-child.github.io/CoreCognition/leaderboard

Metric: Accuracy (%). Source: grow-ai-like-a-child.github.io. Saturation forecast: Estimated already saturated. 33 models tracked. Updated 16 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1InternVL2.5-8B100#2444
2LLaVA-Video-72B-Qwen2100#1228
3VideoLLaMA2-72B100#1355
4qwen2.5-vl-72b-instruct_videoCoreCognition run: native-setting=qwen2.5-vl-72b-instruct_video100
5Llama 3.2 90B Vision Instruct99.17#1769
6InternVL2.5-38B99.17#1413
7VideoLLaMA2.1-7B-16F98.96#1924
8llava-next-video-32B-qwen98.93
9qwen2.5-vl-7b-instruct_videoCoreCognition run: native-setting=qwen2.5-vl-7b-instruct_video98.33
10videollama2-7B-16f98.13
11LLaVA-Video-7B-Qwen297.86#1854
12InternVL2.5-26B97.64#2269
13qwen2.5-vl-3b-instruct_videoCoreCognition run: native-setting=qwen2.5-vl-3b-instruct_video97.5
14yi-vision-v297.22
15llava-video-7B-qwen2-video-only96.17

No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.

Interactive version: theaggregate.ai/benchmark?slug=corecognition-single-image-toolusing · How It Works · Data refreshed daily, snapshot 2026-11-16.