Claude Opus says you can't replace sec.api for my request, are they right?

[penrose-is-goat]

Here is what Claude said about edgartools replacing sec.api:

Can edgartools Replace sec-api.io?
Here's the honest answer: partially, but not as a 1:1 drop-in.
What sec-api.io gives you that's hard to replace for free:

Full-text search across all EDGAR filings by keyword → EFTS can do this, but my implementation was broken
PDF rendering of any filing → sec-api.io converts filings to PDF on demand. edgartools can get filing text directly (via filing.text() or filing.html()), which actually skips the PDF step entirely — this could be better than your current approach since you wouldn't need PyMuPDF at all
The structured response with company name, CIK, ticker, form type, filing URL all in one clean object

[dgunning]

Lots to unpack here. edgartools covers most of the SEC Edgar api and so does sec-api.io.

edgartools is a Python library and sec-api.io is an API provider. There are other differences but most importantly -

What the workflow you are thinking of for comparison between the 2?

[dgunning]

edgartools has full text search via the EFTS endpoint

  from edgar import search_filings

  results = search_filings("artificial intelligence", forms=["10-K"])
  results = search_filings("supply chain risk", ticker="AAPL")
  results = search_filings("tariff impact", forms=["8-K"], start_date="2024-01-01")

  for r in results:
      print(r.form, r.company, r.filed)
      filing = r.get_filing()  # Load the full Filing object

It takes time for documentation updates to make it into model training data