healthcare benchmark
HealthBench Leaderboard
HealthBench is an open-source benchmark for measuring the performance and safety of large language models in healthcare.
Updated Sep 5, 2026
Models9
Model coverage9
MetricScore
EvidenceB
HealthBench Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score60.20% | Percentile100.00% | Participants9 | EvidenceC | Evaluated |
| Rank02 | ModelMA | Score58.00% | Percentile87.50% | Participants9 | EvidenceC | Evaluated |
| Rank03 | ModelOP | Score57.60% | Percentile75.00% | Participants9 | EvidenceC | Evaluated |
| Rank04 | ModelOP | Score57.00% | Percentile62.50% | Participants9 | EvidenceC | Evaluated |
| Rank05 | ModelOP | Score57.00% | Percentile50.00% | Participants9 | EvidenceC | Evaluated |
| Rank06 | ModelOP | Score55.80% | Percentile37.50% | Participants9 | EvidenceC | Evaluated |
| Rank07 | ModelOP | Score54.10% | Percentile25.00% | Participants9 | EvidenceC | Evaluated |
| Rank08 | ModelOP | Score51.40% | Percentile12.50% | Participants9 | EvidenceC | Evaluated |
| Rank09 | ModelOP | Score42.50% | Percentile0.00% | Participants9 | EvidenceC | Evaluated |
HealthBench Highlights
The leading models and scores on this benchmark.
HealthBench Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for HealthBench
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is HealthBench?
What HealthBench measures and how its scores work.
HealthBench consists of 5,000 multi-turn conversations evaluated by 262 physicians using 48,562 unique rubric criteria across health contexts and behavioral dimensions.
HealthBench measures large language model performance and safety in healthcare, reported as a Score ratio.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- HealthBench
- Modality
- text
- Primary category
- healthcare
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about HealthBench.
Which model scores highest on HealthBench?
Qwen3.8 Max is currently ranked first with 60.20%.
What are the top three models on HealthBench?
The current leaders are Qwen3.8 Max (60.20%), Kimi K2-Thinking-0905 (58.00%), and GPT OSS 120B (57.60%).
Which HealthBench model has the lowest official input price?
GPT-5.6 Luna has the lowest matched official input price at $0.20 input / $1.2 output per 1M tokens.
Which models are fastest among HealthBench results?
The fastest matched records are GPT OSS 20B (1,000.00 tok/s via Groq), GPT OSS 120B (500.00 tok/s via Groq), and GPT-5.5 Instant (206.89 tok/s via OpenAI).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.