Image and PDF Uploads

Claude reads uploaded images, PDFs, Word files, spreadsheets and text files on every plan, up to 500 MB per file and 20 files per chat. PDFs up to 100 pages are read with their charts and images; longer ones up to 1,000 pages as text only.

Claude accepts uploads of up to 500 MB per file and up to 20 files per chat, images up to 8000 x 8000 pixels and PDFs up to 1,000 pages, on every plan including the Claude Free Plan (as of Nov 6, 2026). It reads text and visuals in PDFs of 100 pages or fewer and only the text of longer ones. Images can be photos, screenshots, charts, handwriting or diagrams, which every current Claude model can read (Claude Models). What Claude does not do is make images of its own: see Image Generation.

Quick facts

File size (chat) 500 MB per file
Files per chat Up to 20
Image size Up to 8000 x 8000 pixels
PDF length Up to 1,000 pages; visuals read only up to 100 pages
Project files 30 MB per file, no fixed count (Projects)
Document types PDF, DOCX, CSV, TXT, HTML, ODT, RTF, EPUB, JSON, XLSX
Image types JPEG, PNG, GIF, WebP
Plans Every plan, Free included

Source for the table: Upload files to Claude, Claude Help Center (updated Jul 23, 2026), read Nov 6, 2026.

How do I upload a file to Claude?

  1. Click the "+" button at the lower left of the message box.
  2. Choose "Add files or photos" and pick the files.

You can also drag files into the chat window or paste a copied image from the clipboard. Files added to a chat are used in that chat; files added to a project's Files section are available to every chat in the project (Upload files to Claude). With the Google Drive connector, you can also add Google Docs and other Drive files to a chat by URL or from recent files (Claude for Google Workspace).

On iPhone, Anthropic's Oct 29, 2026 app update shows photos in the chat as soon as you pick them, so you can keep typing while they upload in the background (Claude's new mobile app updates are here).

Which file types can Claude read?

Claude takes documents in PDF, DOCX, CSV, TXT, HTML, ODT, RTF, EPUB, JSON and XLSX, and images in JPEG, PNG, GIF and WebP. Excel files (XLSX) need "Code execution and file creation" switched on in Settings > Capabilities, because Claude opens them with code (Upload files to Claude). With that setting on, Claude can also work with other data files and write charts and new files from them (File Creation).

For anything other than a PDF, Claude extracts the text only: images embedded in a Word file, for example, are not read. To have Claude look at a figure inside such a document, save it as a PDF or upload the image on its own.

How does Claude read PDFs?

A PDF of 100 pages or fewer is read with both its text and its visual parts (charts, graphs, images and layout). From 101 to 1,000 pages, Claude reads the text only and ignores the visuals. Above 1,000 pages the upload fails with "Uploaded file is too large" (Upload files to Claude). Splitting a long report into parts of 100 pages or fewer keeps its charts readable.

On the Claude API, the limits are set differently: a request can be up to 128 MB and hold up to 600 PDF pages (100 when the model's context window is under 1M tokens), and each page is read as text plus an image, typically 1,500 to 3,000 tokens a page, with no extra PDF fee (PDF support). Password-protected or encrypted PDFs are not accepted there. Developers can upload a file once with the Files API and reuse it.

What can Claude do with images?

Claude reads what is in an image: it can describe a photo, read text in a screenshot or a scanned page, pull numbers from a chart or table, explain a diagram, follow a handwritten note or check a design mockup. Since Claude Opus 4.7, Anthropic's models accept images up to 2,576 pixels on the long side, more than three times the detail of earlier models; larger images are scaled down before Claude sees them (Vision). Animated GIFs are read from their first frame only.

Anthropic's vision docs list the limits plainly. Claude refuses to name people in images. It can misread low-quality, rotated or very small images (under 200 pixels), its counts of many small objects are approximate, and so are any positions it gives. It cannot tell whether an image is AI-generated. It can look at general medical images but is not designed to interpret diagnostic scans such as CTs or MRIs.

Claude also cannot edit an uploaded photo or return a changed image: it can only describe it, or write code (an SVG, an HTML page) that draws something new (Can Claude produce images?).

Limits worth knowing

  • Two counts of 20. The help center allows up to 20 files per chat, while the API's vision docs describe 20 images per turn on claude.ai. Both were read on Nov 6, 2026.
  • Older 30 MB figure. Earlier help text gave 30 MB per file for chat uploads, and the file creation article still gives 30 MB for files in Claude's code environment (Create and edit files); the upload article, updated Jul 23, 2026, gives 500 MB for chat uploads.
  • Context still counts. Every page and image takes room in the context window and uses part of your usage limit, so a long PDF in a long chat runs out sooner.
  • Project files are text. Files in a project's knowledge are reduced to their text, except PDFs read with their visuals.

Frequently asked questions

What is the Claude file upload limit? Claude accepts files up to 500 MB each and up to 20 files per chat, with images up to 8000 x 8000 pixels and PDFs up to 1,000 pages, according to the Claude Help Center on Nov 6, 2026. Project knowledge files are limited to 30 MB each.

Can Claude read images inside a PDF? Yes, for PDFs of 100 pages or fewer: Claude reads their charts, graphs and images along with the text. In PDFs from 101 to 1,000 pages, Claude reads the text only.

Can the free plan of Claude upload files? Yes. File and image uploads work on the Claude Free plan as on paid plans, though large files use up Free's smaller usage limits quickly.

Can Claude read Excel files? Yes. Claude reads XLSX and CSV files when "Code execution and file creation" is on in Settings > Capabilities. Inside Excel itself, the Claude for Excel add-in works on the open workbook on paid plans.

Can Claude read handwriting? Claude can read handwritten notes in an uploaded image or PDF, though messy handwriting can produce mistakes worth checking.

Can Claude edit or generate images? No. Claude reads and describes images but does not generate or edit photos or illustrations, as of Nov 6, 2026. It can draw diagrams and charts as code instead.

Sources

Moments

Getting started with projects in Claude.ai

Anthropic

  1. You can also drag files into the chat window or paste a copied image from the clipboard. Files added to a chat are…

Claude's new mobile app updates are here

Claude

  1. On iPhone, Anthropic's Oct 29, 2026 app update shows photos in the chat as soon as you pick them, so you can keep…

Getting started with Claude.ai

Anthropic

  1. Claude takes documents in PDF, DOCX, CSV, TXT, HTML, ODT, RTF, EPUB, JSON and XLSX, and images in JPEG, PNG, GIF and…

See also

About this page

Last updated Nov 6, 2026 by the Vi.ki team. Page text CC BY-SA 4.0.

Edit History

Hotkeys

/
Jump to the search box
? or Ctrl + /
Open this list
Esc
Close this list, the photo viewer, or the account menu

Video

Space or K
Play or pause
← →
Back or forward 5 seconds
J L
Back or forward 10 seconds
0 to 9
Go to 0% to 90% of the video
, .
Previous or next frame, while paused
M
Mute or unmute
C
Captions on or off
F
Full screen on or off
A D
Previous or next moment

Photo viewer

← → or A D
Previous or next photo
↑ ↓ or W S
Previous or next photo
Home End
First or last photo
I
Show or hide the photo's info
Z
Clickless on or off: a pointer over a thumbnail shows it
N
Compare with the last photo: back to the one shown before this, and N again comes back
Q
Move the thumbnails to the next edge: top, bottom, left and round
, or Ctrl + ,
Open or close the viewer's settings
V
Clickless pan: a zoomed photo follows the pointer, rather than being dragged
Shift
Browse on or off
B
Show or hide the button bar
T
Show or hide the thumbnails
1
100% zoom on or off
0
Rate the photo as no opinion; again takes the answer back
+ -
Rate the photo a step higher or lower
F
Full screen: the photo alone. F again puts everything back.
Back to top