MMOOC - Partial Answerability (Open-Ended VQA): leaderboard
Metric: Shifted in-context score (%; mean of accuracy on the answerable Partial Answerability Open-Ended VQA questions and answer rationality, the mean 0-1 judge score of the answer's correctness and evidence consistency in steps of 0.25; judge scores averaged over GPT-5.6, Claude Opus 5 and DeepSeek-V4-Pro). Source: arxiv.org. Saturation forecast: Around 2029. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3.5 122B A10B | 72.25 |
| 2 | Gemma 4 31B | 70.75 |
| 3 | Gemma 4 26B | 67.5 |
| 4 | Ministral 3 8B | 64.25 |
| 5 | Qwen 3.5 27B | 63 |
| 6 | GPT-4o | 62.5 |
| 7 | Llama 4 Maverick | 58 |
| 8 | O1 | 56.5 |
| 9 | Ministral 3 14B | 55.5 |
| 10 | InternVL3-8B | 53 |
| 11 | O3 | 44.25 |
| 12 | Claude Opus 4.6 | 32.25 |
| 13 | Gemini 3.1 Pro (Preview) | 22.75 |
Interactive version: theaggregate.ai/benchmark?slug=mmooc-partial-answerability-open-ended-vqa · How It Works · Data refreshed daily, snapshot 2026-09-29.