Skip to content

HealthBench Professional raw score

Raw score on physician-authored clinical consult, documentation, and research conversations.

7 models

Higher is better.

  1. 1Claude Opus 5.5Anthropic77.1
  2. 2Claude Sonnet 5.5Anthropic77.1
  3. 3Claude Opus 5Anthropic73.4
  4. 4GPT-6 AstraOpenAI68.2
  5. 5GPT-6.1 SolOpenAI67.2
  6. 6GPT-6 LunaOpenAI61.2
  7. 7GPT-6 SolOpenAI59.5
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Raw rubric score
Tasks
525 physician-authored conversations
Authors
OpenAI (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.