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

# Export an AI doc, or have it read aloud

> Download an AI doc as a PDF, a PDF of images, Word, JSON, Markdown or a web page, what each format keeps, and how to send the text to text to speech.

You can download an AI doc in six formats, copy its text, or send it to Text to speech to hear it. The PDF and Word files keep your [page design](https://docs.tryhamsa.com/media/ai-docs/page-design.md); Markdown and the web page keep the text and its structure.

## Export a doc

**1. Open the doc**

Open a **Completed** doc from **AI docs**.

**2. Open Export**

Click **Export** at the right of the document bar.

**3. Choose a format**

Click a format. The file is named after the doc's title ("document" when it has none).

![The Export menu open: PDF with its hint, PDF as images, Word (DOCX), JSON, Markdown and Web page (HTML)](https://mintcdn.com/hamsa-452fa79a/fdE-9_eEW1W6yh_s/images/media/ai-docs/export-menu-light.webp?fit=max\&auto=format\&n=fdE-9_eEW1W6yh_s\&q=85\&s=de0fe34bd371dec164985ae6110dc594)

## Formats

| Format | What you get | Keeps |
| - | - | - |
| **PDF** | The browser's print window opens with the pages. Choose "Save as PDF" as the printer. The text can be selected and searched. | Everything: page style, background (colour, design or image, on every page or the first), header with its layout, footer, block colours. |
| **PDF as images** | A PDF downloads at once. Each page is a picture, so the text can't be selected. | The same pages as **PDF**. |
| **Word (DOCX)** | A Word file. | The text, the header's logo, title and subtitle, the footer, a background colour, and a background image that fills every page. Not the page style, a design, the header layout, or First page only. |
| **JSON** | A file with the doc's name, its pages (each page's content as HTML), its language and when it was exported. | The content, page by page. For your own tools. |
| **Markdown** | A `.md` file. | Headings, bold and italic, lists, to-dos, quotes, code, links, tables (the first row as the header), callouts as quotes that start with their emoji, toggles as `<details>`. Not colours, highlight, underline, size, alignment, or the header, footer and background. |
| **Web page (HTML)** | An `.html` file that opens on its own in a browser. | The content with its formatting, the title, and the direction (right to left for an Arabic doc). Not the header, footer or background. |

If an export fails, a message says which one ("The PDF was not exported.") and you can try again.

## Read it aloud

**Read it aloud** sends the doc's text to [Text to speech](https://docs.tryhamsa.com/media/text-to-speech/generate-speech.md) as the script, so you can pick a voice and generate it. It's in the document bar and in Cmd+K (Ctrl+K on Windows).

* Text to speech takes up to 3,000 characters at a time. For a longer doc, "Too long for text to speech" offers to **Use the first 3,000 characters**, or **Cancel**.
* If you were already writing a script in Text to speech, it asks before replacing it: **Replace it** or **Keep my script**.
* The doc is saved before you leave. If the last save failed, you're asked first: **Stay**, or **Leave and read it aloud**.
* The button only shows if you may create text to speech in the project.

Generating speech costs Text to speech credits. See [Text to speech](https://docs.tryhamsa.com/media/text-to-speech/overview.md).

![The Too long for text to speech dialog with the doc's character count, Use the first 3,000 characters and Cancel](https://mintcdn.com/hamsa-452fa79a/fdE-9_eEW1W6yh_s/images/media/ai-docs/read-aloud-too-long-light.webp?fit=max\&auto=format\&n=fdE-9_eEW1W6yh_s\&q=85\&s=eecdf8388286900c6fe076094f6adf8e)

## Copy the text

**Copy text** in the document bar copies the whole doc as plain text, without formatting.

## Good to know

* To check the layout before you export, click **Preview pages**: it shows the doc as the PDF will look, page by page.
* The PDF uses the browser's print window. To get a file, choose "Save as PDF" instead of a printer.

## Related

> [card] [Page design](https://docs.tryhamsa.com/media/ai-docs/page-design.md)
>
> The background, header and footer the PDF and Word files keep.

> [card] [Generate speech](https://docs.tryhamsa.com/media/text-to-speech/generate-speech.md)
>
> Turn the doc's text into speech.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.