Added: PDF Support for Extract Data & Send Prompt

10 Jul 2026
Elizabeth Long
Whippy workflow builder showing a resume-processing flow: Email Parser, Extract Data, Find Candidates, Send Prompt, and Send Email steps, with the Send Email step configured to reply to a job candidate automatically

Whippy added PDF support to the Image field on Extract Data and Send Prompt across the OpenAI, Anthropic, and Gemini apps.

What's included

  • Added PDF support to the Image field on Extract Data and Send Prompt, in addition to images, on all three LLM apps.
  • Added support for PDF URLs, base64, and data URIs, the same input formats already used for images.
  • Added automatic file-type detection, so the workflow identifies and processes PDFs without extra configuration.
  • Kept full backward compatibility: existing workflows that pass images continue to work exactly as before.

How it works

The Image field on an Extract Data or Send Prompt step accepts a PDF the same way it accepts an image. No separate action, no new field, no extra setup. Whippy detects the file type automatically and sends it to the selected provider (OpenAI, Anthropic, or Gemini) in the format that provider expects. If a file isn't a valid PDF or image, the step returns a clear error instead of a silent failure.

Why it matters

Many of the highest-value document workflows, such as resumes, applications, contracts, invoices, injury reports, and intake forms, start as PDFs, not images. Before this update, customers had to convert or OCR those files outside Whippy before a workflow could use them. That step is no longer necessary: the PDF goes straight into Extract Data or Send Prompt on whichever provider the customer already uses, with the same model picker and placeholder support as any other step in Whippy Workflows. Because the update applies to all three providers at once, customers choose a provider based on their own account and preference, not file-type support. For example, a workflow can now extract candidate data directly from a resume PDF without converting it to an image first.

Setup

No new setup required to use this feature. Make sure the app version is up to date: OpenAI 1.0.5 or higher, Anthropic 1.0.4 or higher, and Gemini 1.0.4 or higher. Once updated, existing Extract Data and Send Prompt steps work with PDFs immediately, using the same Image field and the same connection already configured.

Notes

This change affects the Image field on Extract Data and Send Prompt only. No changes were made to how existing image-based workflows behave.