CMDB-1500 (Image): leaderboard
Metric: Accuracy (%), fixed denominator of 300 questions (invalid answers count as wrong), effort as labelled. Source: surdai.com. Saturation forecast: Around January 2027. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6.1 Sol (Low) | 92 |
| 2 | GLM-5.3 Flash (Low) | 90.67 |
| 3 | GPT-6 Luna (Low) | 89.33 |
| 4 | Qwen 3.8 Flash | 83 |
| 5 | Kimi K2.6 | 81.33 |
| 6 | DeepSeek V4.1 Flash | 79.67 |
Interactive version: theaggregate.ai/benchmark?slug=cmdb-1500-image · How It Works · Data refreshed daily, snapshot 2026-10-07.