StacksGPT
Documents & Finance

Extract Data from PDFs with Claude or Gemini

Send a PDF straight to Claude or Gemini to extract and reshape the information you need, and compare the two.

The Problem

Classic PDF extraction chains OCR with a separate model, which loses layout and adds steps.

How This Template Solves It

An operations analyst picks a PDF in Google Drive and writes what to extract in a prompt. The file is sent to both Claude and Gemini in a single step each, so the analyst can compare results before choosing a model.

How It Works

  1. 1Downloads a PDF from Google Drive.
  2. 2Encodes the file so it can be sent inside an API request.
  3. 3Uses your prompt to define what to extract and how to transform it.
  4. 4Sends the PDF and prompt to Claude and to Gemini through HTTP requests.

What You Need

  • A Google Drive account
  • An Anthropic API key, a Gemini API key, or both

What Stays Under Your Control

  • You compare the outputs and choose the model for production use.

Watch Out For

  • Ask for JSON output, as the notes show, when the data feeds another system.
  • Disable one of the API calls if you only want to use one model.

Adapted from a community workflow template. Before publishing we removed credentials, account and file IDs, email addresses and sample data. The original author's setup notes are kept inside the workflow file.

More Documents & Finance Templates

Documents & Finance

Draft RFP Responses from Your Company Knowledge

Extract every question from an RFP document and draft answers from your company documents with an OpenAI assistant.

Starts on: Incoming webhook

GmailGoogle DocsOpenAI+1 more