Find the fillable fields on any form

Today we’re releasing form field detection on the Datalab API. POST /api/v1/form-fields takes a PDF or an image and returns every place on each page where a person is meant to write, tick, or sign. Each field comes back with a type, a bounding box, and the printed label that tells a person what goes there. The endpoint reads the page image, so a scan or a flat PDF works the same way as a fillable one.

Here it is on three public forms. Hover over a box to find its row, or switch to JSON to see the page as the API returns it:

SS-4W-9Form 1040, 1919

• Text 32
• Checkbox / radio 44
• Date 2
• Name / contact 4
• Signature / initials 1

83 fields

Fields JSON

1. company Legal name of entity (or individual) for whom the EIN is being requested
2. company Trade name of business (if different from name on line 1)
3. text Executor, administrator, trustee, “care of” name
4. text Mailing address (room, apt., suite no. and street, or P.O. box)
5. text Street address (if different) (Don’t enter a P.O. box.)
6. text City, state, and ZIP code (if foreign, see instructions)
7. text City, state, and ZIP code (if foreign, see instructions)
8. text County and state where principal business is located
9. name Name of responsible party
10. text SSN, ITIN, or EIN
11. radio Yes
12. radio No
13. radio Yes
14. radio No
15. checkbox Sole proprietor (SSN)
16. text Sole proprietor (SSN)
17. checkbox Estate (SSN of decedent)
18. text Estate (SSN of decedent)
19. checkbox Partnership
20. checkbox Plan administrator (TIN)
21. text Plan administrator (TIN)
22. checkbox Corporation (enter form number to be filed)
23. text Corporation (enter form number to be filed)
24. checkbox Trust (TIN of grantor)
25. text Trust (TIN of grantor)
26. checkbox Personal service corporation
27. checkbox Military/National Guard
28. checkbox State/local government
29. checkbox Church or church-controlled organization
30. checkbox Farmers’ cooperative
31. checkbox Federal government
32. checkbox Other nonprofit organization (specify)
33. text Other nonprofit organization (specify)
34. checkbox REMIC
35. checkbox Indian tribal governments/enter547
36. checkbox Other (specify)
37. text Other (specify)
38. text State
39. text Foreign country
40. checkbox Banking purpose (specify purpose)
41. text Banking purpose (specify purpose)
42. checkbox Shared for applying (check only one box)
43. text Shared for applying (check only one box)
44. checkbox Changed type of organization (specify new type)
45. text Changed type of organization (specify new type)
46. text Shared for applying (check only one box)
47. checkbox Purchased going business
48. checkbox Hired employees (Check the box and see line 13.)
49. checkbox Created a trust (specify type)
50. text Created a trust (specify type)
51. checkbox Compliance with IRS withholding regulations
52. checkbox Created a pension plan (specify type)
53. text Created a pension plan (specify type)
54. checkbox Other (specify)
55. text Other (specify)
56. date Date business started or acquired (month, day, year). See instructions.
57. text Closing month of accounting year
58. text Agricultural
59. text Household
60. text Other
61. checkbox If you expect your employment tax liability to be $1,000 or less in a full cal
62. checkbox Health care & social assistance
63. checkbox Wholesale—agent/broker
64. checkbox Construction
65. checkbox Rental & leasing
66. checkbox Transportation & warehousing
67. checkbox Accommodation & food service
68. checkbox Wholesale—other
69. checkbox Retail
70. checkbox Retail estate
71. checkbox Manufacturing
72. checkbox Finance & insurance
73. checkbox Other (specify)
74. text Indicate principal line of merchandise sold, specific construction work done, pr
75. radio Yes
76. radio No
77. name Designee’s name
78. text Designee’s telephone number (include area code)
79. text Address and ZIP code
80. text Designee’s fax number (include area code)
81. text Applicant’s telephone number (include area code)
82. signature Signature
83. date Date

Source: IRS · output: POST /api/v1/form-fields, no post-processing

Why forms are hard to automate

A fillable PDF stores its fields, but the names are often meaningless to a person. The 2025 Form 1040 names its fields like topmostSubform[0].Page1[0].f1_14[0]. And many forms that reach a pipeline aren’t fillable at all: they’re scans, PDFs exported from Word, or fillable PDFs that someone printed, signed, and scanned back.

Even when a form is fillable, its author decides which blanks become fields. The IRS Form W-7 has no fillable field for either of its signature lines or for the dates beside them.

So the blanks have to be found on the page, and they come in many shapes. The 1913 Form 1040 uses dotted leader lines and splits every amount into dollar and cent columns. The DS-82 asks yes-or-no questions with pairs of boxes. Tax forms print a box for each digit of a Social Security number. Instruction pages have no fields at all, and a detector has to return nothing for them.

What you get back

Every field has one of 12 types:

Type	Used for
text	Any written answer that doesn’t fit a more specific type
checkbox, radio	Tick boxes, and options where only one can be chosen
date	Dates, including a date split into month, day, and year boxes
signature, initials	Places to sign or initial
name, firstName, lastName	Names of people
email, company, title	Contact details, company names, and job titles

The model picks the type from the printed label and the layout around the blank. The second line of the W-9 asks for a business name and comes back as company. The blank after “Date” in the signature row comes back as date.

With the printed label on each field, you don’t have to match a box to nearby text yourself. Fields are listed in reading order. The model is trained to return only empty fields.

Accuracy and speed

In our tests, form field detection finds 87% of form fields, and the median page takes 1.1 seconds. Pages with more fields take longer, because the model writes each field out in turn.

Using it

The endpoint works like our other APIs: submit a file, then poll request_check_url until the status is complete.

import os, time, requests

headers = {"X-Api-Key": os.environ["DATALAB_API_KEY"]}

with open("fw9.pdf", "rb") as f:
    resp = requests.post(
        "https://www.datalab.to/api/v1/form-fields",
        headers=headers,
        files={"file": ("fw9.pdf", f, "application/pdf")},
    ).json()

while True:
    result = requests.get(resp["request_check_url"], headers=headers).json()
    if result["status"] == "complete":
        break
    time.sleep(2)

for page in result["pages"]:
    for field in page["fields"]:
        print(page["page"], field["field_type"], field["label"], field["bbox"])

This is part of the result for page 1 of the W-9, as the model returned it:

{
  "page": 0,
  "fields": [
    {"field_type": "text", "bbox": [0.094, 0.143, 0.941, 0.166], "label": "Name of entity/individual. An entry is required. (For a sole proprietor or d"},
    {"field_type": "company", "bbox": [0.094, 0.176, 0.941, 0.197], "label": "Business name/disregarded entity name, if different from above."},
    {"field_type": "checkbox", "bbox": [0.118, 0.226, 0.131, 0.236], "label": "Individual/sole proprietor"},
    ...
    {"field_type": "signature", "bbox": [0.194, 0.731, 0.623, 0.757], "label": "Signature of U.S. person"},
    {"field_type": "date", "bbox": [0.658, 0.731, 0.941, 0.757], "label": "Date"}
  ],
  "error": ""
}

bbox is [x0, y0, x1, y1] as fractions of the page width and height, with the origin at the top left. To place a field on a PDF page, multiply by the page size in points and flip the y axis, because PDF coordinates start at the bottom left:

from pypdf import PdfReader

reader = PdfReader("fw9.pdf")
for page in result["pages"]:
    box = reader.pages[page["page"]].mediabox
    width, height = float(box.width), float(box.height)
    for field in page["fields"]:
        x0, y0, x1, y1 = field["bbox"]
        rect = (x0 * width, (1 - y1) * height, x1 * width, (1 - y0) * height)
        print(field["field_type"], field["label"], rect)

You can send a file_url instead of uploading the file, limit the pages with page_range (like 0,5-10) or max_pages, and pass a webhook_url to get called when the request finishes. Set processing_location=eu to keep page images in the EU. The endpoint takes PDFs, images, and the other document formats we accept, but not spreadsheets.

If the model can’t read a page, that page comes back with an error message and no fields, and you aren’t charged for it. The request fails only if every page fails.

Where it fits

• Signature workflows: find the signature, initials, and date fields on an uploaded contract or onboarding packet, and place signing tags there automatically.
• Making flat PDFs fillable: write a form field at each box, using the type to pick a text field, a checkbox, or a signature field.
• Digitizing paper forms: build a web form from a paper one, with the printed labels as the questions.
• Before form filling: our /fill endpoint fills a form from data you already have. Field detection tells you what a form asks for, so you can collect those values first.
• Finding blanks a person missed: run it on a submitted form to list what was left blank.

Pricing

Form field detection costs $3 per 1,000 pages.

If you have forms that trip it up, send a few to [email protected].