JMed48k: leaderboard
We introduce JMed48k, a multi-profession Japanese healthcare licensing benchmark for evaluating vision-language models. Built from official PDF materials released by the Japanese Ministry of Health, Labour and Welfare, JMed48k contains 48,862 exam questions and 20,142 images from 11 national licensing examinations between 2005 and 2025, with visual content annotated under an 8-type taxonomy. From
Metric: Overall (self-reported). Source: benchmarklist.com. Saturation forecast: Around May 2027. 21 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 92.8 | #152 |
| 2 | GPT-5 | 92.3 | #98 |
| 3 | GPT-5 Mini | 87.1 | #186 |
| 4 | Claude Sonnet 4 | 84.1 | #199 |
| 5 | Gemini 2.5 Flash | 83.3 | #245 |
| 6 | Gemma 4 31B (IT) | 80.6 | #176 |
| 7 | Grok 4.20 | 77.8 | #175 |
| 8 | Llama 4 Maverick | 73.8 | #464 |
| 9 | MedGemma-4B-IT | 32.3 | #853 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=jmed48k · How It Works · Data refreshed daily, snapshot 2026-10-20.