MedMosaic - Cough Sound MCQ: leaderboard
MedMosaic (Rajgarhia et al., 2026): 3,882 multiple-choice questions on cough recordings (CoughVID, HLS-CMDS) about cough type, timing and counts. Accuracy in percent (higher is better).
Metric: Accuracy (%; multiple choice, extracted answer letter against the answer key; 3,882 questions on cough recordings from CoughVID and HLS-CMDS). Source: arxiv.org. Saturation forecast: Estimated already saturated. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 67 | #262 |
| 2 | Gemini 2.5 Flash | 52.8 | #452 |
| 3 | Qwen2.5-Omni-7B | 48.4 | #1797 |
| 4 | Gemma 3n E4B (IT) | 48 | #2220 |
| 5 | Phi-4 Multimodal Instruct | 47.9 | #3914 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=medmosaic-cough-sound-mcq · How It Works · Data refreshed daily, snapshot 2026-11-12.