MedLoCoMo: leaderboard
Metric: Score (%) over all 17,892 questions: item-weighted combination of judged accuracy on answerable questions (two per admission) and abstention accuracy on unanswerable ones (one per admission); full patient history in context (100 synthetic multi-admission doctor-patient conversations from MIMIC-IV, 74.5k tokens on average), short free-text answers; answerable items judged correct or not by a Gemini 3 Flash Preview judge, adversarial unanswerable items scored by abstention; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.1 | 72.8 |
| 2 | Qwen 3.5 27B | 68.1 |
| 3 | Qwen 3.5 9B | 51.8 |
| 4 | Qwen 3.5 4B | 48.8 |
| 5 | Gemma 3 27B | 40.3 |
| 6 | Gemma 3 12B | 35.7 |
| 7 | MedGemma-4B | 18.4 |
| 8 | Lingshu-32B | 16.5 |
| 9 | Gemma 3 4B | 12.6 |
Interactive version: theaggregate.ai/benchmark?slug=medlocomo · How It Works · Data refreshed daily, snapshot 2026-09-29.