CoreCognition Image + Video Conservation: leaderboard

CoreCognition Image + Video 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. 120 models tracked. Updated 16 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Qwen 2.5 VL 72B Instruct81.68#848
2Qwen-VL-Max-2025-01-2578.98
3O1 (2024-12-17)78.38#313
4eagle-x4-13B-plus75.98
5eagle-x4-8B-plus74.77
6Qwen 2 VL 72B Instruct72.97#1388
7mPLUG-Owl371.47#2436
8H2OVL-Mississippi-2B70.27#5329
9GPT-4o69.97#744
10llama-3-longvila-8b-512framesCoreCognition run: native-setting=llama-3-longvila-8b-512frames67.57
11Kosmos266.97#9796
12Gemini 1.5 Flash-8B66.67#2670
13Qwen 2.5 VL 7B Instruct65.17#1867
14llama-3-longvila-8b-1024framesCoreCognition run: native-setting=llama-3-longvila-8b-1024frames65.17
15LLaVA-Next-Mistral-7B64.86#8137

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

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