Adding to the list — I've been building Deroom AI which generates AI room redesigns from a photo (not text prompts). It preserves the room's geometry and perspective so the output still matches the actual space layout, which I think matters for SketchUp users who want to test material/style ideas on a rendered scene before committing in V-Ray or Enscape.

Free to try (5 credits/day). Curious if anyone here has tried photo-based vs text-based AI viz tools in their workflow.