Articles · Gemini AI Tools
Edit a photo with Gemini and keep the original on the same page
Describe-a-change editing is a second generate, not a layer mask. You upload a JPG, PNG, WebP, or GIF, type what should change, and Gemini returns a new picture. The original stays visible so you can compare. DevNestro does not claim pixel-perfect edges, intact small text, or a reversible history. The same image-generation rate limit applies (three requests per hour by default) because the host uses an image-generation model.
Why this is worth doing in the browser
People ask for background swaps, object removal, added props, style changes, and composition tweaks. Those examples are on the page so the job is clear. The instruction tells the model to apply only the requested change. Models still redraw more than you asked. Keep the original until you accept the result. This is not the local crop or compressor, and it is not a Photoshop replacement.
How to use the tool
Upload the original. Describe the change. Confirm the file may be sent to Google Gemini. Generate edited image. Compare both sides. Download the new preview if you want a copy. Clear removes the original preview and the result from the session.
- Keep the original until you decide the edit is close enough.
- Ask for one change at a time.
- Do not expect tiny text or logos to survive a restyle.
- Skip confidential photos unless you accept Google processing.
Privacy
The uploaded image and the instruction are sent to Google Gemini. DevNestro does not archive either version. Temporary Gemini files are deleted after the reply when a Files API upload was used and cleanup succeeds. This is cloud AI.
Willing to send a photo to Gemini for a draft edit? Open the AI Image Editor, keep the original in view, and download only after you compare.