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

# Supported Formats

> File formats Nvisy can analyze and redact

Every file belongs to a **modality** — `text`, `tabular`, `image`, or `audio` —
which determines how it is analyzed and which redaction operators apply.

## Formats

<Tabs>
  <Tab title="Text">
    | Extension     | Notes                                                    |
    | ------------- | -------------------------------------------------------- |
    | `pdf`         | Born-digital or scanned; scanned pages go through OCR    |
    | `docx`        | Container format — embedded media is analyzed separately |
    | `rtf`         | Rich text                                                |
    | `txt`         | Plain text                                               |
    | `log`         | Plain text                                               |
    | `html`, `htm` | Markup                                                   |
    | `xml`         | Markup                                                   |
    | `json`        | Structured text                                          |
  </Tab>

  <Tab title="Tabular">
    | Extension | Notes                                       |
    | --------- | ------------------------------------------- |
    | `csv`     | Delimited rows                              |
    | `xlsx`    | Container format — sheets analyzed per part |
  </Tab>

  <Tab title="Image">
    | Extension     | Notes                       |
    | ------------- | --------------------------- |
    | `png`         |                             |
    | `jpg`, `jpeg` |                             |
    | `tif`, `tiff` | Multi-page images supported |
  </Tab>

  <Tab title="Audio">
    | Extension | Notes                      |
    | --------- | -------------------------- |
    | `wav`     | Transcribed, then analyzed |
  </Tab>
</Tabs>

<Note>
  Filtering by extension expands to the whole format's extension set — passing
  `jpg` also matches `jpeg`, and `tif` also matches `tiff`.
</Note>

## Container Formats

Formats like `docx` and `xlsx` are containers: they hold parts that may have
their own modality. An image embedded in a Word document is analyzed as an
image, in its own right.

The run's audit reflects this. Each container part appears in `parts`, keyed by
its container-private ID — for a DOCX, that is the zip entry name:

```json theme={null}
{
  "body": { "modality": "text", "entities": [] },
  "parts": {
    "word/media/image1.png": { "modality": "image", "entities": [] }
  }
}
```

This is why a document can produce both text and image findings from a single
run, each redacted with operators appropriate to its modality.

## OCR

Scanned documents have no selectable text layer and must be rendered to images
before recognition. A run's scope controls this:

```json theme={null}
{
  "scope": {
    "ocrMode": { "kind": "force", "dpi": 300 }
  }
}
```

| Mode    | Behavior                                                |
| ------- | ------------------------------------------------------- |
| `auto`  | Extract the text layer, render only pages that lack one |
| `force` | Always render pages to images at the given `dpi`        |
| `never` | Never render; rely on the existing text layer           |

<Warning>
  `auto` is the intended default, but the text-layer parser that drives the
  decision is not in place yet — today only `force` actually renders. Use
  `force` for scanned documents.
</Warning>

## Filtering Files

List endpoints filter by format, modality, or name:

```typescript theme={null}
const page = await client.files.listFiles(workspaceSlug, {
  formats: ["pdf", "docx"],
  modality: ["text"],
  search: "contract",
  limit: 50,
});
```

## File Roles

Each file records the role it plays, which drives retention and whether it is
user-facing:

| Kind       | Description                                  |
| ---------- | -------------------------------------------- |
| `original` | Source document, uploaded or imported        |
| `redacted` | Redacted output produced by a pipeline       |
| `audit`    | Engine analysis blob; hidden from file lists |

Redacted outputs are linked to their source through `parentId`, forming a
version chain rather than overwriting the original.

## Next Steps

<CardGroup cols={2}>
  <Card title="Redaction Workflow" icon="workflow" href="/features/redaction-workflow">
    How modality drives redaction
  </Card>

  <Card title="API Reference" icon="code" href="/api-reference/introduction">
    File endpoints and filters
  </Card>
</CardGroup>