# LiveBench Contamination Audit

> LiveBench scores 8.3/10 (Tier A, Low Risk) on BenchAudit as of November 2026, reflecting a hybrid public and private held-out split and isolated offline third-party grading.

- **Canonical URL:** https://benchaudit.org/benchmarks/livebench/
- **Benchmark ID:** `livebench`
- **Hub:** LiveBench / Abacus.AI
- **Category:** General
- **Domains:** Math, Coding, Science, Business
- **Skills:** Reasoning, Coding, Instruction Following, Agentic, Writing
- **BenchAudit Score:** 8.3/10 (Tier A - Low Risk)
- **Last Audited:** 2026-11-17 (Ref 2026-11)

## Factor Scores

| Factor | Score (/10) | Weight |
| --- | ---: | ---: |
| Privacy | 7.5 | 40% |
| Originality | 7.2 | 25% |
| Freshness | 10 | 20% |
| Grading | 10 | 15% |

## Primary-Source Evidence

### Privacy
> "The current LiveBench release is 2025-04-25; however, not all questions for this release are public on Huggingface. In order to evaluate all categories, you will need to pass --livebench-release-option 2024-11-25 to all scripts to use the most recent public questions."
- Source: https://github.com/LiveBench/LiveBench

### Originality
> "LiveBench is designed to limit potential contamination by releasing new questions monthly, as well as having questions based on recently-released datasets, arXiv papers, news articles, and IMDb movie synopses."
- Source: https://github.com/LiveBench/LiveBench

### Freshness
> "Adds the LiveBench-2026-06-25 release (agentic coding v2) to the old leaderboard site — a new release date, refreshed agentic coding data, and one new model."
- Source: https://github.com/LiveBench/livebench.github.io/pull/45

### Grading
> "# ANSWER-LEAK / NETWORK ISOLATION: run the agent container with no network."
- Source: https://raw.githubusercontent.com/LiveBench/LiveBench/main/livebench/agentic_code_runner/minisweagent/config/livebench.yaml

## Top Frontier Models on LiveBench

| Rank | Model | Creator | Percentile |
| ---: | --- | --- | ---: |
| #1 | [Gemini 4 Neon](https://benchaudit.org/models/gemini-4-neon__xhigh/) | Google | 100.0p |
| #2 | [GPT-6.2 Astra](https://benchaudit.org/models/gpt-6-2-astra__max/) | OpenAI | 100.0p |
| #3 | [GPT-6 Astra](https://benchaudit.org/models/gpt-6-astra__max/) | OpenAI | 100.0p |
| #4 | [Claude Fable 5.2](https://benchaudit.org/models/claude-fable-5-2__max/) | Anthropic | 97.9p |
| #5 | [Claude Fable 5.1](https://benchaudit.org/models/claude-fable-5-1__max/) | Anthropic | 94.5p |
| #6 | [Claude Opus 5.5](https://benchaudit.org/models/claude-opus-5-5__max/) | Anthropic | 91.2p |
| #7 | [Gemini 4 Carbon](https://benchaudit.org/models/gemini-4-carbon__high/) | Google | 87.8p |
| #8 | [GPT-6.1 Sol](https://benchaudit.org/models/gpt-6-1-sol__max/) | OpenAI | 84.5p |
| #9 | [GPT-6.2 Sol](https://benchaudit.org/models/gpt-6-2-sol__max/) | OpenAI | 81.1p |
| #10 | [GPT-6.1 Terra](https://benchaudit.org/models/gpt-6-1-terra__max/) | OpenAI | 77.8p |
