# MMMU-Pro Contamination Audit

> MMMU-Pro scores 2.1/10 (Tier D, Compromised) on BenchAudit as of November 2026, reflecting publicly downloadable plaintext test files and standardized independent third-party grading.

- **Canonical URL:** https://benchaudit.org/benchmarks/mmmu-pro/
- **Benchmark ID:** `mmmu-pro`
- **Hub:** Artificial Analysis / MMMU
- **Category:** General
- **Domains:** Science, Humanities, Math, Medical, Business
- **Skills:** Multimodal, Reasoning, Knowledge Work
- **BenchAudit Score:** 2.1/10 (Tier D - Compromised)
- **Last Audited:** 2026-11-17 (Ref 2026-11)

## Factor Scores

| Factor | Score (/10) | Weight |
| --- | ---: | ---: |
| Privacy | 0 | 40% |
| Originality | 4 | 25% |
| Freshness | 0 | 20% |
| Grading | 7.3 | 15% |

## Primary-Source Evidence

### Privacy
> "An enhanced MMMU benchmark that eliminates shortcuts and guessing strategies to more rigorously test multimodal models across 30 academic disciplines."
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking

### Originality
> "MMMU-Pro rigorously assesses multimodal models' true understanding and reasoning capabilities through a three-step process based on MMMU: (1) filtering out questions answerable by text-only models, (2) augmenting candidate options, and (3) introducing a vision-only input setting where questions are embedded within images."
- Source: https://arxiv.org/abs/2409.02813

### Freshness
> "arXiv:2409.02813 (cs) [Submitted on 4 Sep 2024"
- Source: https://arxiv.org/abs/2409.02813

### Grading
> "Multiple Choice (10 options) Regex extraction, pass@1"
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking

## Top Frontier Models on MMMU-Pro

| Rank | Model | Creator | Percentile |
| ---: | --- | --- | ---: |
| #1 | [GPT-6.2 Astra](https://benchaudit.org/models/gpt-6-2-astra__max/) | OpenAI | 91.4p |
| #2 | [Claude Opus 5.5](https://benchaudit.org/models/claude-opus-5-5__max/) | Anthropic | 87.1p |
| #3 | [GPT-6 Astra](https://benchaudit.org/models/gpt-6-astra__max/) | OpenAI | 82.8p |
| #4 | [GPT-6.1 Sol](https://benchaudit.org/models/gpt-6-1-sol__max/) | OpenAI | 78.5p |
| #5 | [Gemini 3.8 Flash](https://benchaudit.org/models/gemini-3-8-flash__high/) | Google | 74.2p |
| #6 | [Gemini 3.7 Flash](https://benchaudit.org/models/gemini-3-7-flash__high/) | Google | 69.9p |
| #7 | [GPT-6.1 Terra](https://benchaudit.org/models/gpt-6-1-terra__max/) | OpenAI | 65.5p |
| #8 | [Kimi K3.1 Pro](https://benchaudit.org/models/kimi-k3-1-pro__max/) | Kimi | 61.3p |
| #9 | [GPT-6 Sol](https://benchaudit.org/models/gpt-6-sol__max/) | OpenAI | 57.0p |
| #10 | [Qwen3.8 Max (0902)](https://benchaudit.org/models/qwen3-8-max__xhigh/) | Alibaba | 52.6p |
