# MLCR-AA Contamination Audit

> MLCR-AA scores 10/10 (Tier S, Fortified) on BenchAudit as of November 2026, reflecting a strictly private held-out test set and isolated offline third-party grading.

- **Canonical URL:** https://benchaudit.org/benchmarks/mlcr-aa/
- **Benchmark ID:** `mlcr-aa`
- **Hub:** Artificial Analysis / Wisedocs
- **Category:** Science
- **Domains:** Medical
- **Skills:** Long Context, Reasoning, Knowledge Work, Writing
- **BenchAudit Score:** 10/10 (Tier S - Fortified)
- **Last Audited:** 2026-11-17 (Ref 2026-11)

## Factor Scores

| Factor | Score (/10) | Weight |
| --- | ---: | ---: |
| Privacy | 10 | 40% |
| Originality | 10 | 25% |
| Freshness | 10 | 20% |
| Grading | 10 | 15% |

## Primary-Source Evidence

### Privacy
> "Artificial Analysis evaluates a private held-out set of the two hardest tiers (expert-tier clinical synthesis and compound, multi-part reasoning): 60 questions, each run with 3 repeats. This private set is separate from the publicly released dataset"
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking

### Originality
> "Each case is a realistic synthetic medical record of roughly 25,000 to 64,000 tokens. Unless a question sets its own format, we ask models to answer each part in one paragraph and cite the supporting evidence from the record."
- Source: https://artificialanalysis.ai/evaluations/mlcr-aa

### Freshness
> "[2026/08/31] We are excited to release MLCR v1.1 that comes with some improvements to judging calibration and a release with Artificial Analysis"
- Source: https://github.com/Wisedocs-AI/medical-long-context-reasoning

### Grading
> "Responses are scored by a panel of three LLM judges (Claude Opus 5.5, GPT-6.1 Sol, and Grok 4.7, each at medium reasoning effort) against expert-annotated ground truth and the source medical record."
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking

## Top Frontier Models on MLCR-AA

| Rank | Model | Creator | Percentile |
| ---: | --- | --- | ---: |
| #1 | [Claude Fable 5.2](https://benchaudit.org/models/claude-fable-5-2__max/) | Anthropic | 100.0p |
| #2 | [Claude Sonnet 5.5](https://benchaudit.org/models/claude-sonnet-5-5__max/) | Anthropic | 97.6p |
| #2 | [GPT-6 Astra](https://benchaudit.org/models/gpt-6-astra__max/) | OpenAI | 97.6p |
| #4 | [GPT-6.2 Astra](https://benchaudit.org/models/gpt-6-2-astra__max/) | OpenAI | 94.4p |
| #5 | [GPT-6.1 Sol](https://benchaudit.org/models/gpt-6-1-sol__max/) | OpenAI | 91.3p |
| #5 | [Kimi K3.1 Pro](https://benchaudit.org/models/kimi-k3-1-pro__max/) | Kimi | 91.3p |
| #7 | [GPT-6.2 Sol](https://benchaudit.org/models/gpt-6-2-sol__max/) | OpenAI | 88.2p |
| #8 | [Claude Haiku 5.5](https://benchaudit.org/models/claude-haiku-5-5__max/) | Anthropic | 86.0p |
| #9 | [GPT-6 Sol](https://benchaudit.org/models/gpt-6-sol__max/) | OpenAI | 83.9p |
| #10 | [Tencent Hy4](https://benchaudit.org/models/tencent-hy4/) | Tencent | 81.8p |
