HELM MedHELM - MedDialog — leaderboard
Metric: Jury Score. Source: crfm.stanford.edu. 13 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek R1 | 4.44 |
| 2 | GPT-5 Mini (2025-08-07) | 4.44 |
| 3 | GPT-5 (2025-08-07) | 4.37 |
| 4 | O4 Mini (2025-04-16) | 4.23 |
| 5 | O3 Mini (2025-01-31) | 4.23 |
| 6 | Gemini 2.0 Flash | 4.22 |
| 7 | GPT-4o (2024-05-13) | 4.21 |
| 8 | Claude 3.7 Sonnet (20250219) | 4.2 |
| 9 | GPT-4o Mini (2024-07-18) | 4.15 |
| 10 | Claude 3.5 Sonnet (20241022) | 4.11 |
| 11 | Gemini 2.5 Pro (Preview 05-06) | 3.76 |
| 12 | Gemini 1.5 Pro (001) | 3.72 |
Interactive version: theaggregate.ai/benchmark?slug=helm-medhelm-meddialog · How the rankings work · Data refreshed daily, snapshot 2026-07-22.