healthcare benchmark
HealthBench Professional
HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.
Updated Aug 7, 2026
Published models9
Registry coverage9
MetricScore
EvidenceB
HealthBench Professional leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AN | 66.0% | 100.0% | 9 | C | |
| 2 | OP | 60.5% | 87.5% | 9 | C | |
| 3 | AN | 59.8% | 75.0% | 9 | C | |
| 4 | AN | 57.8% | 62.5% | 9 | C | |
| 5 | OP | 57.7% | 50.0% | 9 | C | |
| 6 | AN | 55.8% | 37.5% | 9 | C | |
| 7 | OP | 55.7% | 25.0% | 9 | C | |
| 8 | OP | 38.4% | 12.5% | 9 | C | |
| 9 | MI | 35.0% | 0.0% | 9 | C |
Score distribution
Top published rows on the benchmark's original scale.
HealthBench Professional highlights
The top published results on this benchmark's own scale.
What is HealthBench Professional?
Definition and scoring fields from the benchmark registry.
HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- HealthBench Professional
- Modality
- text
- Primary category
- healthcare
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- healthbench-professional|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about HealthBench Professional.
Which model scores highest on HealthBench Professional?
Claude Fable 5 is currently ranked first with 66.0%.
What does HealthBench Professional measure?
HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
9 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.