MirrorBench - Level 3 (Self-Referential Mirror Recognition): leaderboard
Metric: Average score (mean of task success rate, step-wise improvement ratio, final completion ratio and peak completion ratio, each at most 1; the final completion ratio is negative when the hand ends farther from the mark than it started) at Level 3, where neither the mirror nor the fact that the mark is on the body of the agent itself is mentioned, over both body settings, on MirrorBench's Isaac Sim mirror self-recognition scenes (7 human and robot bodies, 6 hands, 6 marks; the agent moves its hand with six directional actions to touch a mark on its own body seen in a mirror; step budget the shortest path plus 10), temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 0.34 |
| 2 | Gemini 2.5 Flash | 0.24 |
| 3 | GPT-4.1 | 0.11 |
| 4 | GPT-4o | -0.06 |
| 5 | Claude Sonnet 4.5 | -0.15 |
| 6 | GLM-4.5V | -0.29 |
| 7 | Qwen 2.5 VL 7B | -0.31 |
| 8 | GPT-4o Mini | -0.39 |
Interactive version: theaggregate.ai/benchmark?slug=mirrorbench-level-3-self-referential-mirror-recognition · How It Works · Data refreshed daily, snapshot 2026-10-07.