> ## Documentation Index
> Fetch the complete documentation index at: https://openbench.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Knowledge

> Evaluate AI models on knowledge and comprehension tasks

openbench includes a comprehensive suite of knowledge benchmarks covering diverse subjects from undergraduate to graduate level, spanning multiple languages and domains.

## Available Benchmarks

<CardGroup cols={2}>
  <Card title="MMLU" icon="book">
    Massive Multitask Language Understanding across 57 subjects including STEM, humanities, and social sciences.

    ```bash theme={null}
    bench eval mmlu
    ```
  </Card>

  <Card title="MMLU-Pro" icon="graduation-cap">
    Enhanced version of MMLU with more challenging questions and additional subjects.

    ```bash theme={null}
    bench eval mmlu_pro
    ```
  </Card>

  <Card title="GPQA Diamond" icon="gem">
    Graduate-level science questions (PhD-level) in physics, chemistry, and biology.

    ```bash theme={null}
    bench eval gpqa_diamond
    ```
  </Card>

  <Card title="SuperGPQA" icon="star">
    Extended graduate-level question answering spanning 285 academic disciplines.

    ```bash theme={null}
    bench eval supergpqa
    ```
  </Card>

  <Card title="TUMLU" icon="globe">
    Turkish Understanding and Multitask Language Understanding across 9 languages.

    ```bash theme={null}
    bench eval tumlu
    ```
  </Card>

  <Card title="OpenBookQA" icon="book-open">
    Question answering requiring multi-step reasoning with elementary science knowledge.

    ```bash theme={null}
    bench eval openbookqa
    ```
  </Card>

  <Card title="HLE" icon="trophy">
    Humanity's Last Exam - 2,500 expert-written questions from 1,000+ domain experts across diverse fields.

    ```bash theme={null}
    bench eval hle
    ```
  </Card>

  <Card title="HLE Text" icon="file-text">
    Text-only version of Humanity's Last Exam without visual components.

    ```bash theme={null}
    bench eval hle_text
    ```
  </Card>
</CardGroup>

<br />

## Related Resources

* [MMLU Paper](https://arxiv.org/abs/2009.03300)
* [GPQA Paper](https://arxiv.org/abs/2311.12022)
* [OpenBookQA Dataset](https://allenai.org/data/open-book-qa)
