KOFFVQA - Object Attributes: leaderboard
KOFFVQA subscore for recognizing and describing object attributes in Korean open-ended visual questions.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 82 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4.1 | 88.33 | #248 |
| 2 | Gemini 2.5 Pro (Preview 03-25) | 86 | #123 |
| 3 | Gemini 2.5 Flash (Preview 04-17) | 85.83 | #237 |
| 4 | GPT-4.5 (Preview) | 85.5 | #244 |
| 5 | O3 (2025-04-16) | 84.83 | #121 |
| 6 | Gemini 2.5 Pro (Preview 05-06) | 84.67 | #147 |
| 7 | Llama 4 Scout Instruct | 84.33 | #539 |
| 8 | O4 Mini (2025-04-16) | 84.33 | #179 |
| 9 | O1 (2024-12-17) | 83.33 | #149 |
| 10 | GPT-4.1 Mini | 82.5 | #355 |
| 11 | Claude 3.5 Sonnet (20241022) | 81.83 | #291 |
| 12 | Qwen 2.5 VL 72B Instruct | 81.83 | #377 |
| 13 | InternVL3-78B | 81.5 | #352 |
| 14 | Gemini 2.0 Pro (Preview 02-05) | 80 | #307 |
| 15 | Llama 4 Maverick Instruct | 79.17 | #456 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=koffvqa-object-attributes · How It Works · Data refreshed daily, snapshot 2026-10-20.