Eye-Q - Iterative Refinement - Persian: leaderboard
Metric: Exact-Match Accuracy (%; up to three attempts, told when wrong, Persian subset). Source: arxiv.org. Saturation forecast: Around December 2026. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 41.23 |
| 2 | Grok 4.1 Fast (Reasoning) | 24.29 |
| 3 | Gemini 2.5 Flash | 17.84 |
| 4 | GPT-5.2 (Non-reasoning) | 11.09 |
| 5 | Qwen 3 VL 235B A22B Instruct | 4.2 |
| 6 | Llama 4 Scout | 3.77 |
Interactive version: theaggregate.ai/benchmark?slug=eye-q-iterative-refinement-persian · How It Works · Data refreshed daily, snapshot 2026-09-25.