Feature Request: Utilize Vision API for scraping

[mikewolfd]

Offer a node to scrape by taking a full-size image of the website and pass that through to a Vision Inference service.

I've found that this solution results in a more contextually complete response.

[itsperini]

Hey @mikewolfd right now we have only one ImageToText node which uses gpt-4 vision here.
It takes a URL of the image as input. We are planning to implement other multi-modal models such as llava

[VinciGit00]

hi, we have this feature to the screenshot scraper