PDF

Got a manual, a brochure, a whitepaper, a contract template? Upload the PDF — Echo extracts the text and indexes it. Best for formatted documents where the structure matters and copy-pasting into plain text would lose information.

On this page

Why use this

  • You already have a finished, formatted document — manual, brochure, whitepaper, policy, contract template.
  • The content doesn't live on a web page; it's only distributed as a file.
  • You want to preserve the document's structure (headings, tables, ordering) better than a plain text paste would.
  • The document is internal or behind a download, so URL or website crawl isn't an option.

How it works

You select a PDF (up to 10 MB on most plans, unlimited on VIP). Echo uploads it, runs text extraction in the background, and indexes the result. You see a spinner while extraction runs; once it completes, the content appears in your library.

The original file is stored alongside the extracted text. The AI uses the text content to answer; visitors don't typically download the PDF.

How it differs from related options

  • vs Text upload — PDF preserves structure; copy-pasting into text often loses tables, headings, and ordering. If you have a PDF, upload the PDF.
  • vs URL upload — URL is for live web pages. If the PDF lives at a public URL, URL works too, but a direct upload is more reliable for parsing.
  • vs CSV import — CSV is for many small text items at once, not for documents with structure.

Step-by-step setup

  1. Open the assistant and go to Contents.
  2. Click Add content and select PDF.
  3. Pick a PDF file from your device. Maximum 10 MB on most plans; VIP has no size cap.
  4. The file uploads immediately. A spinner shows while text extraction runs.
  5. Once processing completes, the content appears in your library.
Drag and drop
You can drag and drop PDF files directly onto the Contents page. On the unified content library page (top-level Content nav), Echo will ask which assistant should receive the file before uploading.

Common errors and troubleshooting

  • "File too large" — your plan caps PDF uploads at 10 MB. Compress the PDF, split it into smaller files, or upgrade to VIP for unlimited size.
  • "Could not extract text" — scanned PDFs without OCR have no text layer. Run the file through an OCR tool first, or paste the content as text.
  • "Unsupported file type" — only .pdf is supported. Convert Word docs, Pages, or Google Docs to PDF first.
  • Extracted text is messy or missing sections — heavily formatted PDFs (multi-column layouts, watermarks, embedded fonts) can extract poorly. Split the document by section or paste the cleanest version as text.
  • Spinner spins forever — extraction may have failed silently. Refresh the page; if the entry isn't there, re-upload.