Domains: Medical
Skills: Knowledge WorkReasoningWritingInstruction Following

Four-Factor Integrity Breakdown & Primary-Source Evidence

Privacy

Weight 40%
0/10
Access Model: Public Plaintext
"We use the HealthBench and HealthBench Professional datasets that OpenAI publishes with simple-evals (commit 652c89d), pinned by checksum: 5,000 and 525 conversations."

Originality

Weight 25%
10/10
Reuse Profile: All Original (0% reuse)
"Each example includes a physician-authored conversation with ChatGPT for Clinicians and is scored via rubrics written and iteratively adjudicated by three or more physicians across three phases."
Primary source (Paper) ↗ Verified 2026-11-17

Freshness

Weight 20%
3.5/10
Initial Release: 2025-05Latest Refresh: 2026-04Rolling Updates: No
"arXiv:2604.27470 (cs) [Submitted on 30 Apr 2026]"
Primary source (Paper) ↗ Verified 2026-11-17

Grading

Weight 15%
7.3/10
Integrity: Independent StandardizedSandbox: not applicable single turn
"Each model receives the conversation as user and assistant turns, with no system prompt and no tools, and writes the next assistant reply. A judge model checks the reply against each rubric item separately with simple-evals' grader prompt"

Frontier Model Standings on HealthBench

Top 15 active models tracked
Rank Model Official Score Percentile
#1Claude Sonnet 5.5 max Anthropic62.1%100.0p
#2DeepSeek V4.1 Flash max DeepSeek60.9%98.1p
#3Claude Fable 5.2 max Anthropic59.6%93.5p
#4GPT-6.1 Sol max OpenAI58.5%88.9p
#5Claude Haiku 5.5 max Anthropic58.5%84.2p
#6GPT-6 Astra max OpenAI58.2%79.6p
#7Claude Fable 5.1 max Anthropic57.2%74.9p
#8Claude Opus 5.5 max Anthropic57.0%70.2p
#9Grok 4.7 xhigh SpaceXAI56.0%65.6p
#10GPT-6 Luna max OpenAI54.6%61.0p
#11GLM-5.3 max Z AI54.5%56.3p
#12GPT-6 Sol max OpenAI54.2%51.7p
#13GPT-6.2 Sol max OpenAI53.9%47.0p
#14GPT-6.2 Astra max OpenAI53.5%42.4p
#15Gemini 4 Argon high Google49.0%37.8p

Embeddable Audit Badge

View Markdown (.md)

Benchmark maintainers can embed this live SVG badge in GitHub READMEs or documentation:

BenchAudit HealthBench Tier C 4.3/10
[![BenchAudit Tier C](https://benchaudit.org/badges/healthbench.svg)](https://benchaudit.org/benchmarks/healthbench/)