Skip to main content

Index private / paywalled content

Feature Requests

Scenario: I'm logged in into a newpaper site, reading a paywalled article.

  • When asking the AI assistant about the page I'm viewing, it summarizes the paywall login page. It would be cool, if the data visible in the user agent could be sent to OpenAI

  • Similarily, it would be nice, if the full content visible in the client could be indexed, so the article can appear with searches for terms found inside the fenced content.

Posted by [Deleted User] ·
Status: 🎉 Shipped4 comments

Comments4

  • Oliver Sauter

    Team•

    Oct 17, 2023

    Is fixed now too. Goes out with 3.8.5 which is already in review.

    Now you have an option to use the local content if for whatever reason the content can't be fetched on our server side.

    Also cleaned up the mode selector a bit.

  • Oliver Sauter

    Team•

    Oct 4, 2023

    Thanks for leaving this feedback here.
    Indeed we could just take the content on the current page to send it to the summarisation.
    Though it would be much much slower bc we first have to upload the data and then pump it through. Much faster to fetch it client side.

    Wondering if there are good ways of detecting such login/pay walls so we can make it dynamic? Ideas?

    On your second point: That is already the case no matter if they are behind a pay/login wall.
    Did you have an experience where this was not the case?