CoreCognition Single Image Conservation: leaderboard

CoreCognition Single Image Conservation — 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
1qwen2.5-vl-72b-instruct_videoCoreCognition run: native-setting=qwen2.5-vl-72b-instruct_video80.78
2InternVL2.5-78B65.47#1044
3qwen2.5-vl-7b-instruct_videoCoreCognition run: native-setting=qwen2.5-vl-7b-instruct_video65.17
4Llama 3.2 90B Vision Instruct64.56#1769
5yi-vision-v259.16
6InternVL2.5-38B56.76#1413
7qwen2.5-vl-3b-instruct_videoCoreCognition run: native-setting=qwen2.5-vl-3b-instruct_video54.95
8InternVL2.5-8B54.35#2444
9VideoChat-7B54.35
10LLaVA-Video-72B-Qwen254.05#1228
11VideoLLaMA2-72B53.75#1355
12video-chat-13B51.35
13LLaVA-Video-7B-Qwen251.05#1854
14videollama2-7B-16f51.05
15VideoLLaMA2-7B48.05#2213

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-conservation · How It Works · Data refreshed daily, snapshot 2026-11-16.