Pulling FDA Docket Comments per Chat GPT

 Yes—you can automate both the downloading and the screening, without manually opening 129 PDFs. And I found a particularly useful shortcut for this exact docket.

Someone has already collected and analyzed much of it. Innolitics published a “Running Census” of FDA-2026-N-7874. Its author reports retrieving 107 comments and 92 PDF/Word attachments through the Regulations.gov API, then extracting their text. The version I could read describes a September 21 snapshot, so it should not be treated as complete coverage of the 129 you see now. Nevertheless, it provides an immediate starting point for finding substantive comments. Innolitics

For your own complete collection, here are the practical options:

What you want	How to accomplish it
Download everything without clicking each comment	Use the official API to enumerate comments and retrieve each comment’s attachment links; a script downloads the files automatically.
Find comments of three or more pages	Have software count pages in every downloaded PDF and create a sortable index. No manual opening required.
Identify organizations hidden behind personal names	Extract the attachment text and examine letterheads, opening paragraphs, and signatures. Preserve the portal’s submitter name in a separate column.
Search reliably for “California”	Search the locally extracted text of both comment bodies and attachments, using OCR for scanned files.
Prioritize substantive submissions	Combine page count, word count, organizational affiliation, and the presence of concrete recommendations.

The official Regulations.gov API documentation supports retrieving comment details with attachments and explains API-key registration. This is an automated sequence of downloads, rather than necessarily a single “download docket ZIP” button. GSA Open Technology

Your “California” experience is a reason not to trust a zero-result search as proof of absence. One plausible explanation is that the searchable listing contains “John Smith” and “See attached,” while “University of California Health” appears inside the PDF. Attachment indexing, scanning, or posting delays could also matter. I haven’t verified which explanation applies to that particular submission, so I would not categorically claim that the portal never searches attachments.

There is also an Export route, with an important distinction: exporting a spreadsheet of comment records is not the same as downloading their attachments. George Washington University’s Regulatory Studies Center documents an export-plus-script workflow and provides public code for downloading comments and converting attachments to text. Its instructions refer to an older interface/API, so I would use it as a model rather than assume it runs unchanged today. I could not verify the current export-button placement through my web access. GitHub

For your research, the most useful result would be a comment index with:

• Organization identified from the actual letter.
• Portal submitter name and comment ID.
• PDF page count and extracted word count.
• A one-sentence description of the argument.
• Direct attachment link.
• Flags for “three-plus pages,” “needs OCR,” and “duplicate.”

I would use three-plus pages as a priority flag, not an exclusion rule: a two-page institutional letter can contain an important position, while a long attachment may mostly reproduce background material. The automated approach makes both visible.

#### HOW DO IT

####

Absolutely. My previous answer skipped over the difficult part: “use the API” is an instruction for a programmer, not a usable set of directions.

And you’re right that we need the comments submitted near October 19. The September snapshot is insufficient for that purpose. I haven’t verified the submission-date breakdown, so I can’t confirm whether those late comments constitute the majority.

I found an easier place to start: Regulations.gov has a “Bulk Data Download” feature that does not require programming. GSA documents it specifically as a way to download information without technical expertise. It emails you a spreadsheet file. GSA

Here is the plain-English procedure:

1. Open Regulations.gov.
2. Scroll to the bottom of the page and look for “Bulk Data Download.” Alternatively, look on the homepage for “Create Custom Download.”
3. Choose the option to download by Docket ID.
4. Paste this identifier: FDA-2026-N-7874
5. Follow the prompts to request the download and provide your email address.
6. Look for an email from no-reply@regulations.gov. Download the CSV file using its link. “CSV” simply means a spreadsheet that Excel can open.
7. Upload that file here. You don’t need to clean it up first.

Those are GSA’s published directions; the announcement is archived, so the precise button wording or placement may have changed. If your screen differs, send me a screenshot. GSA

Important: that first download is the catalog of comments—not a folder containing all the PDFs. Once you upload it, I can inspect whether it includes the attachment download links. That determines the simplest way to collect the actual files in a batch.

For comparison, here is what the API approach means in ordinary language:

Term	What it means
API	A way for a program to ask Regulations.gov for information automatically.
API key	A registration code that identifies the program’s requests.
Download script	A small program that asks for the comment list, finds each attachment, and downloads the files for you.

The API route requires registering for a key, then running a prepared program. Getting a key alone does not produce the PDFs. The official API supplies comment details and attachment information; the program handles the repeated downloading. GSA Open Technology

For you, I suggest starting with the seven steps above. The immediate goal is to get the current docket’s spreadsheet here. Then we can work toward a folder of PDFs and an index showing organization, page count, and subject—without your opening each submission yourself.