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

# Supported File Types and Size Limits

> Elephas reads your PDFs, Word, PowerPoint, Excel, CSV, Markdown, text, RTF, HTML, EPUB and image files. See every extension, what is read, and size limits.

<Warning>
  Draft. This article depends on an open product question and is not ready to publish.
</Warning>

Elephas reads the everyday files on your desk: PDF, Word, PowerPoint, Excel, CSV, Markdown, plain text, RTF, HTML, EPUB and images, plus web pages by address. Each file can be up to 200 MB, and images up to 50 MB.

## Supported file types

| Type | Extensions | What Elephas reads | Citations point to |
| - | - | - | - |
| PDF | `.pdf` | All text. Scanned pages are read with on-device OCR. | Page |
| Word | `.docx`, `.doc` | Paragraphs, headings and tables. Older `.doc` files are read as plain paragraphs, best effort. | Paragraphs |
| PowerPoint | `.pptx`, `.ppt` | Each slide's title, text, tables and speaker notes. Older `.ppt` files are best effort. | Slide |
| Excel | `.xlsx`, `.xls` | Every sheet, including hidden ones, with column headers kept next to their values. Formulas are read as their last saved values. | Sheet |
| CSV and TSV | `.csv`, `.tsv` | Rows, with the header row kept next to each value | Lines |
| Markdown | `.md`, `.markdown` | Text and headings | Lines |
| Plain text | `.txt` | Text | Lines |
| Rich text | `.rtf` | Paragraphs | Paragraphs |
| HTML | `.html`, `.htm` | The main text and headings. Menus, headers, footers and scripts are left out. | Paragraphs |
| EPUB | `.epub` | Chapters in reading order | Paragraphs |
| Images | `.png`, `.jpg`, `.jpeg`, `.tif`, `.tiff` | Text, read with on-device OCR. Each frame of a multi-page TIFF is a page. | Page |
| Web page | An `http://` or `https://` address | See [add web pages](/docs/brains-and-documents/add-web-pages) | Paragraphs, or page for a PDF |

When you ask, each citation names where the passage came from, such as "page 4", "slide 3", "sheet 2", "paragraphs 12–14" or "lines 40–52". See [citations and sources](/docs/ask/citations-and-sources).

The **Add Files** window only shows supported types. On Windows it may still let you pick another file. Elephas then tells you **This file type is not supported yet.**

## Size and content limits

| Limit | Value | Applies to |
| - | - | - |
| File size | 200 MB | Every file |
| Image size | 50 MB and 50 million pixels | Images |
| Spreadsheet size | 2,000,000 cells per workbook | Excel |
| Scanned pages read with OCR | First 500 pages of a document | Scanned PDFs, multi-page TIFFs |
| Packed content | 5,000 items and 1 GB once unpacked | Word `.docx`, PowerPoint `.pptx`, Excel `.xlsx`, EPUB |
| Reading time | 2 minutes per file | Every file |
| Web page | 20 MB, 30 seconds to load | Web pages |
| Folder | 25,000 files, 8 levels of subfolders | Folders |

A file over the size limit is refused before Elephas copies it: **This file is larger than the import limit.**

A file under the size limit can still hold more than Elephas can safely read, for example a workbook with millions of cells. It then shows **This file has more content than Elephas can safely read.** Only that file is affected. The other files you added carry on.

## Text encoding

Text, CSV, TSV and Markdown files must be UTF-8, or UTF-16 with a byte order mark. Elephas does not guess older encodings. A file in another encoding shows **This file is not valid UTF-8 text.** Open it in a text editor and save it as UTF-8, then add it again.

## Not supported yet

* Apple Pages, Numbers and Keynote files. Export them to PDF, Word, Excel or PowerPoint.
* Apple Notes. Export the notes you need as PDF or text.
* Email files (`.eml`, `.emlx`, `.msg`). Save the emails as PDF.
* HEIC photos, `.rtfd` bundles and Safari `.webarchive` files.
* Audio and video.
* Password-protected Word, Excel and PowerPoint files, and DRM-protected EPUBs. See [password-protected PDFs](/docs/brains-and-documents/password-protected-pdfs) for PDFs.
* Text inside pictures on slides, and in headers, footers, text boxes and footnotes of Word documents.

## Troubleshooting

<AccordionGroup>
  <Accordion title="This file is larger than the import limit">
    Split the file into smaller parts, or for a PDF, save a version with smaller images. Then add the parts.
  </Accordion>

  <Accordion title="This file has more content than Elephas can safely read">
    The file is within the size limit but holds too much content to read safely, such as a very large spreadsheet. Split it into smaller files. For a spreadsheet, export each sheet on its own.
  </Accordion>

  <Accordion title="No readable text was found in this file">
    The file has no text Elephas can extract, for example a PDF made only of drawings. If it is a scan, check that text recognition is installed. See [scanned PDFs and OCR](/docs/brains-and-documents/scanned-pdfs-and-images-ocr) and [file import errors](/docs/troubleshooting/file-import-errors).
  </Accordion>
</AccordionGroup>

## Related articles

<CardGroup cols={2}>
  <Card title="Add files and folders" href="/docs/brains-and-documents/add-files-and-folders" />

  <Card title="Scanned PDFs and OCR" href="/docs/brains-and-documents/scanned-pdfs-and-images-ocr" />

  <Card title="Large libraries" href="/docs/brains-and-documents/performance-and-large-libraries" />

  <Card title="File import errors" href="/docs/troubleshooting/file-import-errors" />
</CardGroup>


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