logoalt Hacker News

lrvicktoday at 3:59 AM0 repliesview on HN

Convert it to an image on the fly to feed it into a vision language model and I expect it would work just fine.