Google appears to be preparing a highly practical upgrade for its Gemini Nano Banana Pro AI model that could significantly enhance how users interact with AI-generated images. According to recent leaks, the tech giant is developing an annotation feature that would allow users to draw directly on images and add text labels before downloading them.
What the Gemini Annotation Feature Offers
The new functionality was revealed by leaker @testingcatalog on X (formerly Twitter), who shared details about Google's development plans. The annotation tool will reportedly include both drawing capabilities and text addition options, represented by a drawing tool and a 'T' icon for text input in the leaked interface.
This feature addresses a growing need among power users who frequently work with AI-generated content. Currently, users have been asking Gemini to annotate images during the generation process, then feeding these annotated images back into the chatbot to create more precise videos using Google's Veo 3.1 video generation technology.
Practical Applications for Content Creators
The annotation feature could revolutionize workflows for digital creators and AI enthusiasts. Users can annotate different parts of images with specific instructions about camera angles, zoom types, or cosmetic changes to make video sequences flow more smoothly and align better with their creative vision.
Another significant application involves providing visual prompts for image generation. Instead of relying solely on text descriptions, users could sketch rough shapes and label them to communicate their ideas more effectively to the AI. For instance, if someone wants to add a cat on the left and a dog on the right of an existing AI-generated image, they could simply draw circles in the desired positions and label them accordingly.
Enhanced Control Over AI Output
This upgrade comes at a crucial time when users are seeking more precise control over AI-generated content. While Gemini Nano Banana Pro has demonstrated improved instruction-following capabilities compared to its predecessors, the image generator can occasionally misinterpret prompts or produce unexpected results.
The annotation feature would enable users to make localized corrections to specific areas of an image that didn't generate according to their instructions, rather than regenerating the entire image from scratch. This could save significant time and computational resources while delivering more accurate results.
The leak suggests that Google continues to actively develop its Gemini AI ecosystem despite recent successful launches. The company introduced both Gemini 3 Pro and Nano Banana Pro models earlier this month, which have received positive feedback from users across various applications.
If implemented, this annotation capability would represent one of the most practical upgrades to Gemini Nano Banana Pro yet, bridging the gap between AI generation and human creative input in a more intuitive way.
