Spidercreator

SpiderCreator is an open source library that uses large language models to help generate web scraping spiders built on Playwright. Its approach is to apply an LLM only while the spider is being created. The resulting spider then runs on ordinary browser automation, so repeated scraping runs avoid per page model costs, which the project argues can be high when an LLM extracts data directly. It is aimed at developers and technical teams with recurring data extraction needs who want to set up scrapers without deep expertise in Playwright or DOM manipulation. The generated extraction logic is kept reviewable and reusable rather than hidden inside a model call. The project says its roots go back to 2019 and to early experiments in AI based web mining and DOM analysis, and that the current version applies that work to Playwright spiders. The project states plainly that the library is highly experimental. Its code, issues and pull requests live on GitHub, contributions are welcomed, and there is a Discord community for discussion. The site does not publish any pricing, and it presents the project as a library to adopt from its repository rather than a hosted service with paid plans.