Document Format Conversion
Upload Word documents, scanned images, Excel files, and more. Instafill.ai converts them automatically so they work as forms to fill or as source data to extract from - no manual export, no reformatting.
Overview
You do not need to convert documents before uploading them to Instafill.ai. Word files, scanned images, Excel spreadsheets, CSVs, and flat PDFs are all accepted directly - the system handles conversion in the background so you can move straight to filling.
Document conversion happens in two contexts. The first is uploading a form template: if you upload a Word document or a scanned image as the form you want to fill, Instafill.ai converts it to a fillable PDF automatically. The second is attaching source data: when you attach any supported file type to a fill session, the system extracts the relevant content from it regardless of format - whether that is a spreadsheet, a Word document, a JPEG of an insurance card, or a scanned handwritten note.
Both paths work the same way from your side: upload or attach the file, and the conversion happens without any extra steps. If you are just getting started and want to see what the platform can do, the ai document filler home page has a full overview.
Supported File Formats
As form templates (the document being filled)
- PDF - fillable PDFs with interactive fields, and flat PDFs with no interactive fields
- Word (.doc, .docx, .docm) - converted to PDF automatically on upload
- Images (PNG, JPEG, TIFF, BMP) - scanned forms converted to PDF with OCR applied
As source data (documents used to extract data from)
- PDF - both text-based and scanned
- Word (.doc, .docx) - text extracted page by page
- Excel and spreadsheets (.xlsx, .xls, .csv, .tsv) - including multi-sheet workbooks
- Images (PNG, JPEG, TIFF, BMP) - with OCR for printed and hand-printed text
- Plain text (.txt) and pasted text input
How It Works
Uploading a Word document or image as a form
When you upload a Word file or image to your Forms dashboard, the conversion starts automatically. A progress indicator appears in your form list while it runs. Once complete, the form status changes to "Ready for Review" and you can check the detected fields before filling.
Word documents are converted to PDF while preserving the layout - paragraph structure, tables, headers, and spacing are retained in the output. The resulting PDF is then processed by the field detection pipeline the same way any uploaded flat PDF would be.
Scanned images go through OCR first to produce a text-searchable PDF. This allows field detection to locate blank areas and labeled sections in the document. For best results on scanned documents, use clean scans at 300 DPI or higher.
Once converted, the template is saved permanently in your account. Every future fill uses the same converted template - the conversion runs only once per document.
For a detailed walkthrough of converting Word and flat PDF documents into fillable forms, see How to create fillable PDF forms automatically with Instafill.ai.
Attaching files as source data
In any fill session, the Files tab accepts any supported format. Upload a spreadsheet with client data, attach a Word-format intake form, or add a scanned insurance card - the AI reads all of them and extracts the relevant data for filling.
The system handles each format appropriately. Spreadsheet data is read row by row and column by column. Word documents are read page by page. Image files go through OCR. All extracted content is then mapped to the target form's fields using the same AI pipeline regardless of the source format.
Multiple source files of different types can be attached in the same session. A fill session might include a Word document, an Excel spreadsheet, and a scanned ID - the AI combines all three sources and fills the form from the combined data.
For more detail on how source documents are used in filling, see Autofill from Multiple Sources.
Scanned documents and OCR
Scanned paper forms, photographs of documents, and image-based PDFs all go through optical character recognition before field detection runs. This produces a text layer on top of the visual image, which the field detection pipeline uses to identify where fields are and what each one is labeled.
Accuracy on scanned source documents is highest for clearly printed text. Hand-printed block letters are recognized well at good scan quality. Cursive handwriting is supported on a best-effort basis and may require more manual field review after filling.
Use Cases
Converting a Word form template
Legal teams, HR departments, and healthcare providers often maintain form templates as Word documents. Uploading a .docx directly - an employment agreement, a patient intake form, a compliance checklist - converts it to a fillable PDF without any manual export. The template is saved once and reused for every future fill session. See how this workflow plays out at scale in a real estate law firm flat PDF automation case study.
Processing scanned paper forms
Healthcare providers, government agencies, and insurance companies frequently deal with paper forms that have been scanned to TIFF or JPEG. Uploading these images converts them to searchable PDFs, which can then be processed by the flat-to-fillable conversion pipeline to create fillable templates for future use.
Using spreadsheets as source data
A batch fill session using a CSV or Excel file converts the spreadsheet data into form fills automatically. Each row becomes one completed PDF. The AI maps column headers to form field labels, handles date formatting, checkbox logic, and value normalization - all from a standard spreadsheet export with no preprocessing required.
Attaching mixed-format source documents
Immigration attorneys receive supporting documents in many formats - a birth certificate scanned as JPEG, an employment letter as Word, a financial statement as PDF. All three can be attached as sources in a single fill session. The system reads each format and extracts the relevant data for filling the target form.
Key Capabilities
| Capability | Detail |
|---|---|
| Word-to-PDF | .doc, .docx, and .docm files converted automatically on upload, layout preserved |
| Image-to-PDF | PNG, JPEG, TIFF (multi-page), and BMP files converted with OCR applied |
| Spreadsheet as source | .xlsx, .xls, .csv, and .tsv files read as source data in fill sessions and batch jobs |
| OCR on scanned documents | Printed and hand-printed text recognized in scanned images and image-based PDFs |
| One-time conversion | Templates are converted once and saved permanently - no re-conversion on future fills |
| Mixed-format sessions | Attach Word, PDF, image, and spreadsheet sources together in a single fill session |
| Automatic format detection | The system identifies the file type and routes it to the correct conversion path - no manual selection |
| Batch source processing | CSV and Excel files used in batch filling are processed row by row at scale |
Benefits
- Upload the file you already have - no need to export from Word, convert manually, or reformat before uploading
- One conversion per template - the result is saved permanently and reused for every future fill
- Mixed-format source files work in the same session - attach a spreadsheet, a Word doc, and a scanned image together and the AI reads all three
- Scanned documents get OCR applied automatically, making field detection possible without manual text entry
- Spreadsheet data from any standard export (Excel, CSV, TSV) works directly as source input or batch fill data
Common Questions
Which file formats can I upload as a form template?
PDF (fillable or flat), Word (.doc, .docx, .docm), and image files (PNG, JPEG, TIFF, BMP). Word and image files are converted to PDF automatically when you upload them. The resulting PDF is then processed for field detection and saved as your reusable template.
Which file formats can I attach as source data in a fill session?
PDF, Word (.doc, .docx), Excel (.xlsx, .xls), CSV, TSV, images (PNG, JPEG, TIFF, BMP), and plain text files. You can attach multiple files of different types in the same session. You can also paste text directly into the Input tab.
Does Word-to-PDF conversion preserve the document's layout?
Yes. Paragraph structure, tables, headers, footers, and embedded images are preserved in the converted PDF output. For form templates where field position matters, review the detected fields in the review screen after conversion and adjust any that did not align correctly with the document's layout.
How does OCR work on scanned documents?
When you upload a scanned image or an image-based PDF, the system applies optical character recognition to produce a text layer on the document. This text layer allows field detection to identify labeled sections and blank areas in the scanned form. For best results, use clean scans at 300 DPI or higher. The accuracy is highest for clearly printed text.
Can I use a scanned handwritten form?
Scanned handwritten documents can be uploaded and processed. Printed and clearly hand-printed block text is recognized well. Cursive handwriting is supported on a best-effort basis - recognized content may need more manual review after filling. The scanned image itself is always preserved accurately as the visual record of the document.
Do I need to convert my Word document to PDF before uploading?
No. Upload the .doc or .docx file directly. The conversion happens automatically. If you are using the standalone Create Fillable PDF tool, you can also upload Word files directly there for conversion with additional controls for confidence, resolution, and page selection.
Does converting a Word document produce fillable PDF fields?
Not automatically. A Word document converted to PDF produces a flat PDF - the visual layout is preserved but there are no interactive form fields. The flat PDF then goes through the field detection step, which identifies where fields should be based on the document's layout (blank lines, underscores, labeled spaces) and creates fillable fields in those positions. You review the detected fields before saving the template.
Can I use Excel or CSV files in batch filling?
Yes. Batch filling is built around spreadsheet input. Upload a .csv, .xlsx, or .xls file where each row represents one form fill, and the system generates one completed PDF per row. Column headers are mapped to form field labels automatically. See Batch Processing for the full workflow.