After introducing native image editing earlier this year, Google is giving the Gemini app capability a major upgrade thanks to “nano banana.”
That’s how Google has been testing and teasing its latest image editing model in recent weeks. Officially, it’s Gemini 2.5 Flash Image for those in Google AI Studio and the Gemini API, and currently the top-rated Image Edit model on LMArena.
In the Gemini app, it’s integrated into whatever model you’re using. Since April, the improvement Google DeepMind has been focusing on is “maintaining a character’s likeness from one image to the next.”
…our latest update is designed to make photos of your friends, family and even your pets look consistently like themselves, whether you’re trying out a 60’s beehive haircut or putting a tutu on your chihuahua.
An example character consistency prompt is: “Reimagine this person as a… matador inside a bullfighting ring, artist, or 90s sitcom.”
Meanwhile, you can blend photos together by uploading multiple and asking the prompt to combine: “Create an image where the woman in the photo is cuddling the dog on a basketball court.”
Similarly, you can “apply the style of one image to an object in another” with design mixing:
- Turn this into a stunning dress on a woman walking down a street in New York.
- Reimagine these rain boots. The shape and style completely inspired by the flowers image.
Finally, multi-turn editing remains, with Google’s demo showing how you can visualize what an empty room would look like with a new coat of paint and specified furniture.
Gemini’s upgraded image editing capabilities are rolling out starting today. They feature an “ai” watermark and the invisible SynthID.
More on Gemini:
- Google rolls out new pill-shaped Gemini overlay on Android
- Gemini Live on Pixel 10 gets native audio, while Pixel Watch 4 lets you raise-to-talk
- Gemini app rolls out search and nav drawer redesign, direct copying on Android
FTC: We use income earning auto affiliate links. More.
Comments