Eye-Q - Iterative Refinement - Arabic: leaderboard
Eye-Q (Najar et al., 2026), Arabic visual word puzzles (50 synthetic scenes written by the authors): each image encodes a hidden word or phrase through visual cues, and the model must output the exact target. The model may revise twice after being told its answer is wrong, and the puzzle counts if any of the three attempts is right. Exact-match accuracy in percent; higher is better.
Metric: Exact-Match Accuracy (%; up to three attempts, told when wrong, Arabic subset). Source: arxiv.org. Saturation forecast: Around June 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 17.02 | #262 |
| 2 | Gemini 2.5 Flash | 8.51 | #452 |
| 3 | Grok 4.1 Fast (Reasoning) | 8.51 | #426 (Grok 4.1 Fast) |
| 4 | Qwen 3 VL 235B A22B Instruct | 4.26 | #521 |
| 5 | GPT-5.2 (Non-reasoning) | 4.26 | #172 (GPT-5.2) |
| 6 | Llama 4 Scout | 0 | #1787 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=eye-q-iterative-refinement-arabic · How It Works · Data refreshed daily, snapshot 2026-11-12.