Best AI models for perceptual illusions
Ranked by accuracy on the meo “perceptual illusions” domain — one of eleven un-leaked, contamination-resistant domains. Per-domain strength reveals where a model genuinely reasons versus where it pattern-matches; the leaders differ sharply by domain.
Ranking 12 models across the meo-tested roster · as of 2026-06-08.
| # | Model | Lab | accuracy |
|---|---|---|---|
| 1 | OpenAI: GPT-5.5 | openai | 100% |
| 2 | Google: Gemini 3.5 Flash | 80% | |
| 3 | Anthropic: Claude Opus 4.8 | anthropic | 78% |
| 4 | Google: Gemini 3.1 Pro Preview | 74% | |
| 5 | Perceptron: Perceptron Mk1 | perceptron | 56% |
| 6 | MiniMax: MiniMax M3 | minimax | 55% |
| 7 | StepFun: Step 3.7 Flash | stepfun | 55% |
| 8 | Qwen: Qwen3.7 Plus | qwen | 54% |
| 9 | xAI: Grok 4.3 | x-ai | 53% |
| 10 | Google: Gemini 3.1 Flash Lite | 53% | |
| 11 | Xiaomi: MiMo-V2.5 | xiaomi | 52% |
| 12 | MoonshotAI: Kimi K2.6 | moonshotai | 51% |