Gemini Gets a ChatGPT Competitor for Image Editing

Users of the Gemini app and developers on Google's AI Studio and Vertex AI platforms will get access to the upgraded image-editing tool. This new model, known internally as "nano-banana," enables complex, text-prompted photo edits.

The update's main advantage is its high precision and ability to preserve details. Asking most current AI services to recolor parts of an image can lead to disappointing results—faces might warp, and backgrounds can shift unexpectedly. Gemini 2.5 Flash Image handles these tasks effectively, working correctly with fine details like facial features, animal fur texture, or patterns on clothing.

Google also states it has implemented strict guardrails into the system. The AI is prohibited from generating intimate imagery of people without their consent. All generated images are automatically tagged with special metadata and watermarks so users can distinguish them from real photographs.

The launch of this precise image editor is a significant move for Google in its competition with OpenAI, whose ChatGPT bot sees more weekly users than Gemini has in a month. The new feature in Gemini will be useful for everyday tasks, like visualizing renovation ideas or trying on clothing styles, as the model can composite multiple items into a single image based on an example. Full access for all users is set to arrive by the end of August.

The expanded functionality for the image editor follows the recent expansion of NotebookLM's availability, which gained support for over 80 languages. It seems this fall is shaping up to bring a wave of new innovations from Google's AI team.