Gemini Updates Editing Capabilities for Images

• Gemini offers detailed image editing using natural language.
• Users can perform edits step by step and combine photos.
• The 'nano banana' model ensures the preservation of the authenticity of objects during editing.

This is reported by Mezha, ZDNet.

In mid-2025, Google introduced a new update for its Gemini app, which significantly improves the capabilities for editing images. Now users can edit photographs in detail, utilizing natural language for formulating queries.

One of the key features is the preservation of the authenticity of people and animals during editing. Thanks to the new capabilities, the bot can change the font, combine photographs, or apply other changes, while keeping the original appearance of the objects intact.

Users can apply changes step by step. For example, they can start by cropping a section of the photo, then add stickers, and finally insert their own images. Gemini also allows for the application of different styles to various elements within a single image.

The new image editor is available free of charge in the Gemini app. All images created or edited in Gemini will bear a visible watermarked signature, as well as an invisible digital watermark SynthID, confirming that they were generated by artificial intelligence.

Among the new features is also the 'nano banana' model, developed by DeepMind, which allows for editing images while preserving the essence of the objects. Users can change clothing, location, or combine elements from different photos, creating new compositions.

Tags: Technology/AI