Adobe's Experimental Camera App Will Now Critique Your Photos Like a Pro
Project Indigo uses AI to tell you what's wrong with your shot, what to remove from the frame, and how to reshoot it. Here's what that looks like in practice.
Key points
- Adobe's Project Indigo, an experimental iOS camera app, added AI-powered photo critique and editing suggestions in a 2025 update.
- The app uses Google's Gemini Nano, a compact AI language model, to generate feedback on framing, lighting and emotional impact.
- A new object-removal tool lets users toggle out people, vehicles, wires and clutter without drawing on the screen.
- Adobe says features are available only to select testers and may never reach a general release.
Adobe has built a critic directly into the camera, and it will tell you exactly why your shot is weak.
Project Indigo, Adobe's experimental iOS camera app first reported by TechCrunch, launched last year with professional manual controls and multi-frame super-resolution (a technique that combines several rapid shots into one sharper image). As we reported on 20 July, the app was already leaning on Google's AI rather than Adobe's own. This update doubles down on that, adding large language model features, the same category of AI behind chatbots like ChatGPT, to review photos and suggest improvements.
Marc Levoy, the engineer leading Project Indigo who previously shaped the camera system inside Google's Pixel phones, has a clear philosophy. He argues that asking users to type a text prompt to edit a photo is asking too much, because finding the right words is genuinely hard. So most of the new tools work as buttons and toggles, producing consistent results rather than a creative lottery.
The photo critique feature reads a finished image and returns a short opinion on four things: framing, lighting, colour and emotional impact. It reads like notes from a photography tutor. You might not agree with every call, but the feedback is specific enough to actually learn from.
Capture and edit suggestions go a step further. The feature looks at what's in your viewfinder before you shoot and recommends changes: shift the angle, adjust exposure, move that object on the left. TechCrunch's reviewer was told to remove a hexagonal white object cluttering the frame. A second section then points to specific Adobe Lightroom controls, the sliders inside Adobe's popular photo-editing software, that would improve an existing shot.
Object removal is where things get visually impressive. Apple Photos, Google Photos and Adobe Photoshop have all offered this for years, but they require manually circling whatever you want gone. Project Indigo gives you a checklist instead: background people, wires and poles, fences, vehicles. Tap a toggle, and the AI fills the gap. Early tests show clean results with no strange repeated textures, and the app also let the TechCrunch reviewer remove a friend and the object he was holding without visible artifacts.
Should ordinary photographers actually use this?
Right now, no, because access is limited to selected testers. Adobe is treating this as a live experiment, and these features may never ship publicly. But the direction matters. A camera that explains why a shot is weak is genuinely useful for anyone trying to improve, and the toggle-based removal is a real convenience over manual selection.
The update also includes depth-of-field simulation (a blurred-background effect applied after the shot), style transfer to watercolour or pen-and-ink looks, and a free-text prompt for edits the preset buttons don't cover. Adobe is currently running the AI features on Google's Gemini Nano model but says it could swap in its own Firefly model later. Worth watching: whether Adobe ever ships these tools to the wider Lightroom audience, or keeps Indigo a permanent experiment.
