KOFFVQA - Korean Recognition: leaderboard
KOFFVQA Korean recognition subscore for visual questions requiring Korean-language answers or labels.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 82 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O1 (2024-12-17) | 100 | #149 |
| 2 | Gemini 2.5 Pro (Preview 03-25) | 90 | #123 |
| 3 | GPT-4o (2024-08-06) | 90 | #336 |
| 4 | O3 (2025-04-16) | 90 | #121 |
| 5 | Gemini 2.0 Pro (Preview 02-05) | 90 | #307 |
| 6 | Gemini 2.5 Pro (Preview 05-06) | 90 | #147 |
| 7 | Gemini 2.5 Flash (Preview 05-20) | 90 | #216 |
| 8 | GPT-4.1 | 85 | #248 |
| 9 | GPT-4o (2024-11-20) | 85 | #381 |
| 10 | Gemini 2.5 Flash (Preview 04-17) | 85 | #237 |
| 11 | Claude 3.5 Sonnet (20241022) | 80 | #291 |
| 12 | GPT-4.5 (Preview) | 80 | #244 |
| 13 | O4 Mini (2025-04-16) | 70 | #179 |
| 14 | Gemini 2.0 Flash (Preview) | 70 | #370 |
| 15 | Gemini 2.0 Flash (001) | 70 | #360 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=koffvqa-korean-recognition · How It Works · Data refreshed daily, snapshot 2026-10-20.