Claude AI can now analyze PDFs – here’s how to try it (and why you should)

Anthropic has added a new feature to its Claude 3.5 Sonnet artificial intelligence (AI) model: visual PDFs or the ability to view and analyze PDFs (OpenAI feature). ChatGPT already supports). The model can not only interpret text in a PDF file, but also imagescharts and graphs.

There’s a catch, however: this feature is only available through a paid professional subscription or API. If you’re not a subscriber yet, here’s why you might want to upgrade.

Also: Anthropic warns of AI disaster if governments don’t act within 18 months

According to the Anthropic user manual, the new Claude 3.5 Sonnet works with any standard PDFwith some restrictions. The PDF file cannot be larger than 32 MB or more than 100 pages, and cannot be encrypted or password protected. To help Claude better analyze and understand the PDF, make sure the text is clear and legible, uses standard fonts, and has the correct page orientation.

You can ask Claude questions about any of the text, images, charts, and tables in the PDFs you download. Some examples given by Anthropic include:

• Analyzing financial statements and the charts or tables they contain.
• Extracting key data from legal documents
• Translation of documents into another language
• Convert documents into more structured and organized formats.

The Anthropic User Guide explains how to use the new PDF parsing in the Messages API. This is fine for programmers and other people creating applications using Claude AI, but what about the rest of us? This is where a subscription comes in handy.

Also: How the 2024 US presidential election will shape the future of technology

For $20 a month Pro plan gives you early access to new features among other benefits. And in this case, PDF analysis is a new feature that is currently in beta mode.

How to access the PDF feature of Claude 3.5 Sonnet

If you have a Pro plan, click your account name or email address at the bottom of the left sidebar and select Feature Preview from the menu. In the Feature Preview window, select Visual PDFs and turn its toggle on.

Return to the home screen and start a new chat. Click the paperclip icon and select the PDF file on your computer. Once the file thumbnail appears below the invitation, enter and submit your question or request.

Also: I tested 7 content detectors with artificial intelligence – they became much better at detecting plagiarism.

You can start by asking Claude to do something simple like summarize a file. You can then move on to ask specific questions about the text in the file.

From here, try asking him about a specific image, table or chart. To do this, refer to the page number containing the image or diagram. When doing this, use the logical page number (the number displayed by the PDF viewer) rather than the physical page number (the number displayed on the page). For example, in Adobe Reader, hover over the physical page number on the right and the logical page number will appear.

If the PDF is too large for Claude, you can split it into smaller files and download each one individually. On the Anthropic website GitHub pageyou will also find several example PDFs that can be downloaded and sent to Claude for analysis and responses.