Skip to main content

Benchmark profile

HealthBench Hard

A harder subset of OpenAI's HealthBench for evaluating open-ended medical and health reasoning with rubric-based grading.

Data verified

Benchmark score on HealthBench Hard — July 23, 2026

BenchLM mirrors the published score view for HealthBench Hard. Muse Spark leads the public snapshot at 42.8% , followed by GPT-5.4 (40.1%) and GPT-5.6 Sol (33.1%). BenchLM does not use these results to rank models overall.

8 modelsKnowledgeCurrentDisplay onlyUpdated July 23, 2026

Benchmark score table (8 models)

Score
1
Muse SparkMeta · Closed
42.8%
2
GPT-5.4OpenAI · Closed
40.1%
3
GPT-5.6 SolOpenAI · Closed
33.1%
4
GPT-5.6 TerraOpenAI · Closed
32.7%
5
GPT-5.6 LunaOpenAI · Closed
32.0%
6
Gemini 3.1 ProGoogle · Closed
20.6%
7
Grok 4.20xAI · Closed
20.3%
8
Claude Opus 4.6Anthropic · Closed
14.8%

The published HealthBench Hard snapshot places Muse Spark first at 42.8%. The third row is 9.7 points behind. The broader top-10 range is 28.0 points, so the table still separates the published systems.

8 models have been evaluated on HealthBench Hard. The benchmark falls in the Knowledge category. This category carries a 12% weight in BenchLM.ai's overall scoring system. HealthBench Hard is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About HealthBench Hard

Year

2026

Tasks

1,000 health prompts

Format

Open-ended health evaluation

Difficulty

Advanced health reasoning

Meta describes HealthBench Hard as a 1,000-prompt subset of OpenAI's HealthBench, graded with the same simple-evals implementation and a GPT-4.1-based judge. BenchLM treats it as a display-only health benchmark reference.

BenchLM freshness & provenance

Version

HealthBench Hard 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

FAQ

What does HealthBench Hard measure?

A harder subset of OpenAI's HealthBench for evaluating open-ended medical and health reasoning with rubric-based grading.

Which model scores highest on HealthBench Hard?

Muse Spark by Meta currently leads with a score of 42.8% on HealthBench Hard.

How many models are evaluated on HealthBench Hard?

8 AI models have been evaluated on HealthBench Hard on BenchLM.

Last updated: July 23, 2026 · BenchLM version HealthBench Hard 2026

Choose a model with this week’s evidence

Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.

One email each week. Unsubscribe anytime.