CoreCognition Image only Permanence: leaderboard
CoreCognition Image only Permanence — 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: Rough model projection: around 2026. 77 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | points-qwen-2.5-7B-chat | 100 | |
| 2 | Step-1.5V-mini | 93.75 | #2720 |
| 3 | step-1v-32k | 87.5 | |
| 4 | omchat | 87.5 | |
| 5 | GLM-4V-9B | 84.38 | #3333 |
| 6 | GLM-4V | 78.13 | #3924 |
| 7 | cogvlm2-llama3-chat-19B | 78.13 | |
| 8 | GLM-4V-Plus | 75 | #1400 |
| 9 | Cambrian-13B | 75 | #7053 |
| 10 | wemm-chat-2k-cn | 75 | |
| 11 | Molmo 7B-O | 71.88 | #6979 |
| 12 | Molmo-7B-D-0924 | 71.88 | #7502 |
| 13 | Molmo-72B-0924 | 71.88 | #5272 |
| 14 | step-1o-turbo-vision | 71.88 | |
| 15 | LLaVA-v1.5-13B | 68.75 | #7610 |
Interactive version: theaggregate.ai/benchmark?slug=corecognition-image-only-permanence · How It Works · Data refreshed daily, snapshot 2026-11-16.