Comment by culi

I don't get this attitude. Unless you're just feeding the scraped data into an LLM or doing archival work, you will need to structure the data anyways, right? So either you're gonna do website-specific work to structure the data or you can just get already-structured data from an API. The vast majority of APIs also follow a spec like OpenAPI or standard idioms as well so it's much less repeated work