BenchLeader
Menu
Best LLMsChoose a modelCoding plansLeaderboardsModelsBenchmarksProvidersReleases8

HealthBench Professional (AA)

As of 29 Oct 2026, the top score on HealthBench Professional (AA) is 66.0% by GPT-6 Astra (OpenAI), ahead of GPT-6.1 Sol at 65.1%, across 14 model configurations with a published result (OpenAI, result dated 3 Sept 2026). OpenAI's HealthBench Professional: 525 clinician conversations covering consults, writing and research, graded against physician-written rubrics. Run by Artificial Analysis. HealthBench Professional (AA) is shown for reference only and does not count toward the BenchLeader Index.

Key facts (as of 29 Oct 2026)

  • GPT-6 Astra holds the top HealthBench Professional (AA) score at 66.0% (OpenAI, 3 Sept 2026).
  • GPT-6.1 Sol is second on HealthBench Professional (AA) at 65.1% (OpenAI, 29 Sept 2026).
  • HealthBench Professional (AA): OpenAI's HealthBench Professional: 525 clinician conversations covering consults, writing and research, graded against physician-written rubrics. Run by Artificial Analysis (OpenAI).
  • 14 model configurations have a published HealthBench Professional (AA) result on BenchLeader (BenchLeader, 29 Oct 2026).
  • HealthBench Professional (AA) does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).

OpenAI's HealthBench Professional: 525 clinician conversations covering consults, writing and research, graded against physician-written rubrics. Run by Artificial Analysis.

Published by
OpenAIdata via Artificial Analysis
Category
Knowledge
Index weight
Reference only
Models
14
Data as of
29 Oct 2026
Newest result
7 Oct 2026

Source: Artificial Analysis (artificialanalysis.ai). Data taken from the public leaderboard.

Top 15
  1. 1GPT-6 Astra (max)66.0%
  2. 2GPT-6.1 Sol (max)65.1%
  3. 3Claude Sonnet 5.5 (max)63.0%
  4. 4GPT-6 Sol (max)62.3%
  5. 5GPT-6 Luna (max)61.9%
  6. 6Claude Opus 5.5 (max)59.9%
  7. 7Claude Haiku 5.5 (max)58.4%
  8. 8Claude Fable 5.1 (max)57.8%
  9. 9Grok 4.7 (xhigh)55.8%
  10. 10Gemini 4 Argon (high)55.0%
  11. 11Kimi K3 (max)53.2%
  12. 12GLM 5.3 (max)52.8%
  13. 13Gemini 3.8 Flash (high)51.7%
  14. 14DeepSeek V4.1 Flash (max)50.7%
14 of 14
#
1GPT-6 Astramax66.0%67.72026-09-03
2GPT-6.1 Solmax65.1%67.22026-09-29
3Claude Sonnet 5.5max63.0%65.02026-09-28
4GPT-6 Solmax62.3%63.62026-09-22
5GPT-6 Lunamax61.9%56.62026-09-22
6Claude Opus 5.5max59.9%69.22026-09-22
7Claude Haiku 5.5max58.4%56.72026-10-07
8Claude Fable 5.1max57.8%67.02026-09-01
9Grok 4.7xhigh55.8%58.12026-09-21
10Gemini 4 Argonhigh55.0%65.92026-09-30
11Kimi K3max53.2%62.22026-07-16
12GLM 5.3max52.8%58.92026-08-18
13Gemini 3.8 Flashhigh51.7%61.92026-09-02
14DeepSeek V4.1 Flashmax50.7%58.92026-09-10

More knowledge benchmarks · All LLM benchmarks

Related benchmarks

Top head-to-heads on HealthBench Professional (AA)

Cite as: BenchLeader, “HealthBench Professional (AA) leaderboard”, data as of 29 Oct 2026, https://www.benchleader.com/benchmarks/aa_healthbench_professional.

HealthBench Professional (AA): questions

What does HealthBench Professional (AA) measure?
OpenAI's HealthBench Professional: 525 clinician conversations covering consults, writing and research, graded against physician-written rubrics. Run by Artificial Analysis. Scores are reported in percent of tasks solved; higher is better.
What is the top score on HealthBench Professional (AA)?
As of 29 Oct 2026, the highest score on HealthBench Professional (AA) is 66.0% by GPT-6 Astra (result dated 3 Sept 2026), ahead of GPT-6.1 Sol at 65.1%. Scores come from OpenAI and are re-read daily.
Which AI model leads HealthBench Professional (AA)?
GPT-6 Astra leads HealthBench Professional (AA) with 66.0% as of 29 Oct 2026, ahead of GPT-6.1 Sol at 65.1% (result dated 3 Sept 2026).
How many models have HealthBench Professional (AA) results?
14 model configurations have a HealthBench Professional (AA) result on BenchLeader, all taken from OpenAI via Artificial Analysis.
Who runs HealthBench Professional (AA) and how often is it updated?
HealthBench Professional (AA) is published by OpenAI. BenchLeader re-reads the published results once a day and records the date each result was published.
Does HealthBench Professional (AA) count toward the BenchLeader Index?
No. HealthBench Professional (AA) is shown for reference but left out of the composite index.