KOFFVQA - Document Understanding: leaderboard
KOFFVQA subscore for understanding document-like images and Korean text-rich visual inputs.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 82 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 100 | #123 |
| 2 | O4 Mini (2025-04-16) | 100 | #179 |
| 3 | Gemini 2.5 Pro (Preview 05-06) | 100 | #147 |
| 4 | Gemini 2.5 Flash (Preview 05-20) | 100 | #216 |
| 5 | Gemini 2.5 Flash (Preview 04-17) | 98.67 | #237 |
| 6 | Gemini 1.5 Pro (002) | 94.67 | #349 |
| 7 | GPT-4.1 Mini | 93.33 | #355 |
| 8 | Gemini 2.0 Flash (Preview) | 93.33 | #370 |
| 9 | Gemini 1.5 Flash-8B (001) | 91.33 | #791 |
| 10 | GPT-4.5 (Preview) | 90.67 | #244 |
| 11 | O3 (2025-04-16) | 90 | #121 |
| 12 | O1 (2024-12-17) | 90 | #149 |
| 13 | Gemini 2.0 Flash (001) | 90 | #360 |
| 14 | Gemini 2.0 Flash Lite (001) | 90 | #544 |
| 15 | Gemini 1.5 Flash (002) | 89.33 | #583 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=koffvqa-document-understanding · How It Works · Data refreshed daily, snapshot 2026-10-20.