RapidCloud Chat can now find images by meaning and by visual similarity, not just by text. Ask for “the login screen of the community portal” and Chat surfaces the actual screenshot — rendered inline, right next to the cited text that explains it.
Under the hood it uses CLIP-based semantic search across your image store, plus a visual nearest-neighbour lookup to find more like any result. Everything stays grounded: the images Chat shows are the ones it actually used to answer.
It’s part of our push to make every kind of content — documents, transcripts, and now images — first-class in one grounded conversation.