LAD-Bench - Unprompted Detection: leaderboard
LAD-Bench (2026) tests whether vision-language models notice logical anomalies in everyday scenes, such as a mailbox in the middle of a road or a dolphin with gills, across 1,007 synthetic images; a grading model checks each free-text answer against the annotated anomaly. This board is the percentage of images whose anomaly the model points out when shown the image with no text prompt at all; higher is better.
Metric: Level 1 accuracy (%): share of images whose anomaly the model points out when given the image alone, with no text prompt or hint; LAD-Bench: 1,007 synthetic images (gpt-image-1) with one logical anomaly each (nature, residential, urban and collaborative scenes); free-text answers graded by GPT-5-nano against the annotated anomaly; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 54.22 | #144 |
| 2 | Claude Sonnet 4.6 | 45.58 | #161 |
| 3 | GPT-5 | 44.79 | #167 |
| 4 | GPT-5 Mini | 35.05 | #348 |
| 5 | Grok 4.1 Fast (Reasoning) | 28.6 | #430 (Grok 4.1 Fast) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=lad-bench-unprompted-detection · How It Works · Data refreshed daily, snapshot 2026-11-13.