# HealthBench Professional — Mercor evaluation (HealthBench Professional (Mercor))

> Answering clinicians in care consults, documentation and medical research conversations. This table shows Mercor-run configurations for reference and is excluded from model rankings.

Canonical page: https://benchlm.ai/benchmarks/mercor-healthbench-professional

- Category: [Knowledge](/knowledge)
- Last updated: November 12, 2026 capture

Evaluation results published by [Mercor](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/). Published evaluation aggregates only; task contents and dataset license grants are not included.

We captured Mercor's HealthBench Professional results on November 12, 2026. The table preserves 37 published configurations, their source identifiers, reasoning settings, and disclosed adjusted score values. These are results from Mercor's evaluation setup; we did not rerun them.

This dated source table is display only and does not enter overall or category rankings. Source configurations retain their external identity, including Pro descriptors and different effort settings.

The headline is the adjusted score Mercor publishes for the 525 physician-authored tasks. Margins are not reported. Scores labs report on their own runs, and other evaluators' runs, are listed separately and are not combined with this table.

The row labels retain published error margins and sample counts when supplied. Missing margins and counts remain unreported; a sample count is not used to reconstruct a task denominator. Except where the dedicated source defines an interval, the confidence level is unreported. Equal displayed scores and overlapping margins do not establish statistical ties. Capture time does not establish individual evaluation dates.

The general methodology lists expert-authored rubrics, scored criterion by criterion and a judge of LLM judge over expert rubrics. Its run-count label is k = 4. The linked result page takes precedence where its setup differs, as noted above.

- [Mercor HealthBench Professional results](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/)
- [Mercor APEX benchmark index](https://www.mercor.com/apex/)
- [Mercor methodology](https://www.mercor.com/apex/methodology/)
- [Source paper](https://arxiv.org/abs/2505.08775)
- [Upstream evaluation code](https://github.com/openai/simple-evals)
- [Upstream dataset](https://huggingface.co/datasets/openai/healthbench)

## About HealthBench Professional (Mercor)

- Tasks: 525 clinician tasks
- Format: Adjusted Score
- Difficulty: Source-specific evaluation
- Paper: [HealthBench Professional evaluation results](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/)

Mercor runs the 525 physician-authored HealthBench Professional tasks and publishes the adjusted score. Scores the labs report themselves stay on the HealthBench Professional page.

HealthBench Professional (Mercor) is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (37 configurations)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [GPT-6 Astra](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 64.50% |
| 2 | [GPT-6.1 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 64.40% |
| 3 | [Sonnet 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 62.70% |
| 4 | [GPT-6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 62.20% |
| 5 | [GPT-6 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 61.40% |
| 6 | [GPT-5.6 Sol (Pro)](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 61.40% |
| 7 | [Opus 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 60.40% |
| 8 | [Fable 5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 59.00% |
| 9 | [Haiku 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 58.80% |
| 10 | [Muse Spark Open](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · xhigh reasoning · Margin not reported · 1,050 reported samples | Meta | 57.20% |
| 11 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 56.50% |
| 12 | [GPT-5.6 Terra](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 56.30% |
| 13 | [Grok 4.7](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · xhigh reasoning · Margin not reported · 2,100 reported samples | xAI | 56.20% |
| 14 | [GPT-5.6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | OpenAI | 55.90% |
| 15 | [Opus 5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 55.40% |
| 16 | [Gemini 4 Argon](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 55.40% |
| 17 | [GLM-5.3](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Zhipu | 52.90% |
| 18 | [Sonnet 5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 52.50% |
| 19 | [Gemini 3.8 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 52.40% |
| 20 | [GPT-5.5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · xhigh reasoning · Margin not reported · 2,100 reported samples | OpenAI | 51.20% |
| 21 | [Kimi K3](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Kimi | 50.20% |
| 22 | [DeepSeek-V4.1-Flash](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | DeepSeek | 49.90% |
| 23 | [Opus 4.7](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Anthropic | 49.70% |
| 24 | [Gemini 3.7 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 48.40% |
| 25 | [Gemini 3.6 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 46.40% |
| 26 | [Grok 4.6](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · xhigh reasoning · Margin not reported · 2,100 reported samples | xAI | 44.40% |
| 27 | [Gemini 3.1 Pro](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 43.90% |
| 28 | [Gemini 3.5 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 43.40% |
| 29 | [Grok 4.5](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | xAI | 42.50% |
| 30 | [Beam](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | Reflection | 41.70% |
| 31 | [Inkling](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Thinking Machines | 41.20% |
| 32 | [DeepSeek-V4-Pro-0813](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | DeepSeek | 40.90% |
| 33 | [Qwen3.8-Max](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · xhigh reasoning · Margin not reported · 2,100 reported samples | Alibaba | 39.00% |
| 34 | [MiniMax-M3](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | MiniMax | 37.10% |
| 35 | [Gemini 3.5 Flash Lite](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | Google | 35.00% |
| 36 | [Nemotron 3 Ultra](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · high reasoning · Margin not reported · 2,100 reported samples | NVIDIA | 32.90% |
| 37 | [DeepSeek-V4-Flash](https://www.mercor.com/apex/oss-benchmarks/oss-healthbench/) | Single-shot agent · max reasoning · Margin not reported · 2,100 reported samples | DeepSeek | 32.80% |

## FAQ

### What does HealthBench Professional (Mercor) measure?

Answering clinicians in care consults, documentation and medical research conversations. This table shows Mercor-run configurations for reference and is excluded from model rankings.

### Which model leads the published HealthBench Professional (Mercor) snapshot?

GPT-6 Astra currently leads the published HealthBench Professional (Mercor) snapshot with a score of 64.50%.

### How many models are evaluated on HealthBench Professional (Mercor)?

The November 12, 2026 capture contains 37 source configurations.

### Does HealthBench Professional (Mercor) affect BenchLM's overall score?

Not directly. HealthBench Professional (Mercor) is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.