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
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Qwen 2.5 VL 72B Instruct | 81.68 | #848 | |
| 2 | Qwen-VL-Max-2025-01-25 | 78.98 | ||
| 3 | O1 (2024-12-17) | 78.38 | #313 | |
| 4 | eagle-x4-13B-plus | 75.98 | ||
| 5 | eagle-x4-8B-plus | 74.77 | ||
| 6 | Qwen 2 VL 72B Instruct | 72.97 | #1388 | |
| 7 | mPLUG-Owl3 | 71.47 | #2436 | |
| 8 | H2OVL-Mississippi-2B | 70.27 | #5329 | |
| 9 | GPT-4o | 69.97 | #744 | |
| 10 | llama-3-longvila-8b-512frames | CoreCognition run: native-setting=llama-3-longvila-8b-512frames | 67.57 | |
| 11 | Kosmos2 | 66.97 | #9796 | |
| 12 | Gemini 1.5 Flash-8B | 66.67 | #2670 | |
| 13 | Qwen 2.5 VL 7B Instruct | 65.17 | #1867 | |
| 14 | llama-3-longvila-8b-1024frames | CoreCognition run: native-setting=llama-3-longvila-8b-1024frames | 65.17 | |
| 15 | LLaVA-Next-Mistral-7B | 64.86 | #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.