Rebecca Soskin Hicks et al.

HealthBench Professional

An open benchmark for clinician-facing model responses across care consult, writing and documentation, and medical research tasks.

In index
Knowledge
Best in knowledge
Source ↗

Models18

Top result69.2%Claude Sonnet 5.5

Top-3 spread3.9 pts

Year2026

Result and price

16 of 18 results

Results

18 results

#ModelResult

01Claude Sonnet 5.569.2%

02Claude Opus 5.565.6%

03Ling 3.1 Flash65.3%

04GPT-6 Astra64.7%

05GPT-6.1 Sol64.2%

06GPT-6.1 Terra62.7%

07Muse Wonder Full62.3%

08GPT-6 Sol60.8%

08GPT-6 Luna60.8%

10Muse Wonder60.5%

10GPT-5.6 Sol60.5%

12Claude Opus 559.8%

13Muse Spark 1.159.3%

14GPT-5.6 Terra57.7%

15Grok 4.756.7%

16GPT-5.6 Luna55.7%

17Ling 3.0 Flash Sante54.3%

18GPT-5.448.1%