Mirror, Mirror on the Wall: leaderboard
In the animal kingdom, mirror self-recognition is a canonical probe of higher-order cognition, emerging only in some species. We ask whether an analogous functional capability emerges in embodied vision-language model (VLM) agents: can they recognize themselves in a mirror? We introduce a controlled 3D benchmark where a first-person VLM agent must infer a hidden body attribute from its reflection
Metric: TSA (self-reported). Source: benchmarklist.com. Saturation forecast: Around May 2027. 44 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.6 | 91 | #90 |
| 2 | Gemma 4 26B | 57 | #292 |
| 3 | GPT-5.1 | 52 | #135 |
| 4 | Ministral 3 14B | 33 | #650 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mirror-mirror-on-the-wall · How It Works · Data refreshed daily, snapshot 2026-10-20.