Reference
PDFs from automation platforms, and what each does with the file
Every platform here can make an HTTP request, so that is not the interesting question. The interesting question is what it does with the bytes that come back, and the answers are genuinely different: one hands you a file object that every downstream action accepts, one has no concept of binary at all and needs a URL instead, one is a real runtime where the question does not arise, and one saves the file to storage that is public by default.
Several of these integrations are built on storing the document and passing a reference rather than the file, which is the storage capability and its per-plan retention.
Workflow automation
- Power Automate
Microsoft's cloud flow automation, which sits closest to SharePoint, OneDrive, Outlook and Dataverse and is usually chosen because the documents have to land in one of them.
Code-first automation
- Pipedream
A workflow platform where the steps are real Node.js or Python, so anything the language can do the workflow can do.
- Google Apps Script
JavaScript running inside a Google account, with direct access to Drive, Gmail, Sheets and Docs and no infrastructure to run.
App builders
- Retool
An internal tool builder where the UI is assembled from components and the data comes from queries, which makes a file an awkward shape to pass around.
- Bubble
A visual application builder with its own database and workflow engine, where an external call is configured in the API Connector rather than written.
Databases and spreadsheets
- Airtable
A database with a spreadsheet interface, where attachments are a first-class field type and are set in a way that surprises people.
Platforms with their own guide
These three already have a full walkthrough on this site. A second page each would compete with the first for the same reader, so they are linked rather than repeated.
- n8n
The HTTP Request node returns a file into the item's binary data, which every downstream node that takes a file will accept.
- Zapier
Webhooks has no real binary handling, so the working shape is to store the document and give Zapier the URL, which its file fields fetch for themselves.
- Make
The HTTP module returns file data that Drive, Dropbox and email modules consume directly, as long as response parsing is left off.