Use a PDF as context in any text field
Keep the PDF open in one tab, write in another. ShroomPen reads the PDF and uses it for the answer you are writing.
Published
The short answer
With ShroomPen, a free Chrome extension, you can use a PDF open in one tab as context for text you write in another tab: a form answer, an email reply or a support message. You pick the PDF tab as a source, ShroomPen extracts its text inside the extension, and your AI model (your own key or a local model) writes a draft you insert into the field. The PDF is only read when you pick it, and nothing is saved unless you choose to.
How to do it
Open the PDF in a Chrome tab
Open the PDF in Chrome as usual, from a website over HTTPS or from a file on your computer.
First time? Install ShroomPen and connect a model with the provider setup guide.
Go to the field you want to fill
Switch to the tab with the form, email or text box and click into the field.
Pick the PDF tab as a source
Press Ctrl+Q to open ShroomPen, click Tabs and select the PDF tab. Tabs that look like PDFs are marked PDF.
Allow access when Chrome asks
The first time, Chrome asks you to let ShroomPen reach the site the PDF came from. For a PDF on your computer, turn on "Allow access to file URLs" for ShroomPen and allow local files once.
If "Allow access to file URLs" is off, ShroomPen opens its Chrome settings page so you can turn it on. Chrome restarts the extension when you do, so open ShroomPen again afterwards.
Ask and insert
Type what you need, for example "Answer the eligibility question using section 3 of the PDF", press Enter, check the draft and choose Insert draft.
What it is good for
- Answering application or grant form questions from the call for proposals PDF.
- Replying to a client email with numbers from a quote, invoice or contract.
- Filling a supplier or security questionnaire from your policy document.
- Writing a support reply from a product manual.
What ShroomPen does with the PDF
| Step | What happens |
|---|---|
| Listing tabs | Only titles and addresses are shown. Nothing is downloaded. |
| You select the PDF tab | ShroomPen downloads the PDF from the tab's own address (HTTPS, localhost or a local file), checks it is a real PDF under 25 MB, and parses it with pdf.js inside the extension. |
| Text extraction | Up to 40,000 characters or 80 pages. Table rows are kept as cells separated by "|"; the layout is approximate. |
| You generate | The text is sent with your instruction to the provider you connected, marked as a PDF so the model knows the layout is rough. |
| Afterwards | The text is not written to storage, unless you save the tab to a Workspace. |
Limits
- No OCR: scanned or image-only PDFs yield no text.
- Plain HTTP PDFs from other sites are not read. HTTPS, localhost and local files are.
- Very long PDFs are cut at 40,000 characters or 80 pages. Point the model at the section you need in your instruction.
- A long PDF adds tokens to every request that includes it. On a small paid model this is still well under a cent per answer. See the cost breakdown.
Related pages
Questions
Can an AI Chrome extension fill a form from a PDF?
Yes. With ShroomPen you open the PDF in one tab, click into a form field in another, pick the PDF tab as context and ask for the answer. ShroomPen writes a draft for that field, and you check it and insert it. It fills one field at a time, not the whole form automatically.
Is the PDF uploaded anywhere?
The PDF file is read inside the extension with a bundled copy of pdf.js, so the file itself is not uploaded for parsing. The extracted text is sent to your AI provider only when that tab is selected for your request. With a local model, it stays on your computer.
Does it work with scanned PDFs?
No. ShroomPen reads the text layer of a PDF and has no OCR. A scanned, image-only PDF has no text to read and is reported as unavailable.
How big a PDF can it read?
Files up to 25 MB. It uses up to 40,000 characters or 80 pages of text, whichever comes first.
Can it read a PDF inside Gmail or Google Drive's preview?
It reads PDFs shown in Chrome's own PDF viewer. Previews drawn by a website, like Drive's viewer, are web pages rather than PDF tabs, so open or download the PDF so it shows in Chrome's viewer.