# MKA (Multimodal Knowledge Agent) Contamination Audit

> MKA (Multimodal Knowledge Agent) scores 9/10 (Tier S, Fortified) on BenchAudit as of November 2026, reflecting a strictly private held-out test set and self-reported or live-network evaluation.

- **Canonical URL:** https://benchaudit.org/benchmarks/mka/
- **Benchmark ID:** `mka`
- **Hub:** Artificial Analysis / Google DeepMind
- **Category:** Knowledge Work
- **Domains:** Business, Finance, Legal, Medical, Science
- **Skills:** Multimodal, Agentic, Tool Use, Knowledge Work, Reasoning, Writing
- **BenchAudit Score:** 9/10 (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 | 3.3 | 15% |

## Primary-Source Evidence

### Privacy
> "Evaluation set: We run a 200-task private set across 10 industry verticals. Each task has a prompt, a set of input files (PDFs, Office documents, spreadsheets and images), and a rubric written by domain experts."
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking

### Originality
> "Artificial Analysis' implementation of Google DeepMind's Multimodal Knowledge Agent (MKA) benchmark. It tests whether agents can read visually dense project files, such as architectural drawings, engineering schematics and clinical scans, and produce the deliverables a professional would."
- Source: https://artificialanalysis.ai/evaluations

### Freshness
> "November 10, 2026 Multimodal Knowledge Agent (MKA) benchmark: testing knowledge work capabilities with complex visual files"
- Source: https://artificialanalysis.ai/articles/google-deepmind-multimodal-knowledge-agent

### Grading
> "The agent works in a Linux sandbox that holds the input files and has internet access. It has a code execution tool, web search and page-fetch tools, and an image-viewing tool that we enable only for models that support image inputs."
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking

## Top Frontier Models on MKA (Multimodal Knowledge Agent)

| Rank | Model | Creator | Percentile |
| ---: | --- | --- | ---: |
| #1 | [Claude Fable 5.2](https://benchaudit.org/models/claude-fable-5-2__max/) | Anthropic | 100.0p |
| #2 | [Gemini 4 Neon](https://benchaudit.org/models/gemini-4-neon__xhigh/) | Google | 100.0p |
| #3 | [Gemini 4 Carbon](https://benchaudit.org/models/gemini-4-carbon__high/) | Google | 99.6p |
| #4 | [Claude Opus 5.5](https://benchaudit.org/models/claude-opus-5-5__max/) | Anthropic | 95.6p |
| #5 | [GPT-6.2 Astra](https://benchaudit.org/models/gpt-6-2-astra__max/) | OpenAI | 91.5p |
| #6 | [Claude Sonnet 5.5](https://benchaudit.org/models/claude-sonnet-5-5__max/) | Anthropic | 87.4p |
| #7 | [Gemini 4 Argon](https://benchaudit.org/models/gemini-4-argon__high/) | Google | 83.3p |
| #8 | [GPT-6 Astra](https://benchaudit.org/models/gpt-6-astra__max/) | OpenAI | 79.2p |
| #9 | [GPT-6.1 Sol](https://benchaudit.org/models/gpt-6-1-sol__max/) | OpenAI | 75.1p |
| #10 | [Muse Wonder](https://benchaudit.org/models/muse-wonder__max/) | Meta | 71.0p |
