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

# Multimodal modules

> Moxus AI multimodal modules cover image understanding, image generation, PDF files, video understanding, and video generation.

Moxus AI brings common multimodal workflows into one API gateway. You can call text, image, video, and document-processing models with the same account, API key, billing, and usage logs.

<Info>
  Multimodal availability depends on the selected model, account group, and API key restrictions. Check model capability labels in the model catalog before calling a model.
</Info>

<Columns cols={2}>
  <Card title="Image understanding" icon="image" href="/en/overview/modules/image-understanding">
    Read images for descriptions, recognition, comparisons, extraction, and visual QA.
  </Card>

  <Card title="Image generation" icon="wand-sparkles" href="/en/overview/modules/image-generation">
    Generate images from prompts or edit existing images with supported models.
  </Card>

  <Card title="PDF files" icon="file-text" href="/en/overview/modules/pdf-files">
    Use documents for summarization, QA, field extraction, and structured output.
  </Card>

  <Card title="Video understanding" icon="video" href="/en/overview/modules/video-understanding">
    Analyze scenes, actions, subtitles, frame changes, and clip-level content.
  </Card>

  <Card title="Video generation" icon="film" href="/en/overview/modules/video-generation">
    Run text-to-video, image-to-video, and creative video variant tasks.
  </Card>
</Columns>

## How to choose a module

Start from the input type. Use image understanding for pictures and screenshots, PDF files for long documents, and video modules when the task requires temporal context. When you need predictable downstream data, combine the module with structured output.

## Next steps

* [Vision and image generation](/en/guide/vision-and-image)
* [Embeddings and reranking](/en/guide/embeddings-and-rerank)