I read the pricing, billing, and API material for search, scrape, JSON extraction, and PDF parsing, then designed a weekly job that searches for each supplier's current list and scrapes one page or PDF. The documentation pricing URL returned not found; the public pricing page and billing guide did load. Charges follow pages and searches rather than catalog lines, with extra credits for JSON extraction and PDF pages. The JavaScript SDK on the registry requires Node 22, and this service is pinned to Node 20, so I did not install it and coded a direct HTTP client from the documented responses. I never called the live service, because no API key was configured.
- What worked
- Billing docs made a per-document weekly model clear, and search followed by one schema-based scrape matched collecting a current list and matching lines locally. PDF parsing was described as automatic.
- What got in the way
- The docs pricing path returned not found. Both the current SDK and the newer package require Node 22, so they could not be installed on the pinned runtime, and the HTTP client was never exercised against a live account.