Google announced the integration of Nano Banana AI image generation model into Google Search and NotebookLM on October 13 2025 expanding access to Gemini 2.5 Flash capabilities for image creation and editing. In Google Search users access the tool through Lens and AI Mode by tapping the Create tab to upload photos generate images from text prompts or apply transformations. NotebookLM incorporates Nano Banana into Video Overviews enhancing summaries of uploaded documents with six visual styles and two video formats.
The model has generated over five billion images since August 2025 maintaining subject consistency and detail across edits. Google Photos will receive Nano Banana support for photo memories and editing in upcoming updates. The rollout begins in English for the United States and India on Android and iOS with additional languages and countries planned.
Table of Contents
How does it work in Google Search and NotebookLM
In Google Search Nano Banana activates through the Lens app by selecting the Create tab identified by a banana icon and capturing or uploading an image. Users enter text prompts to transform subjects such as changing outfits hairstyles or backgrounds while preserving facial details and layout. The tool supports standalone generation from descriptions like creating a photo booth image or scene illustration. Generated images include a Gemini spark watermark for provenance and allow downloads sharing or further refinements via follow up prompts. Integration with AI Mode enables prompt based creation without images starting with a Describe your image hint. The model processes requests with low latency supporting resolutions up to 1024×1024 pixels and various styles.
NotebookLM uses Nano Banana to generate contextual visuals for Video Overviews transforming uploaded documents into narrated videos. Users select sources generate the overview and customize with six styles: Watercolor Papercraft Anime Whiteboard Retro Print and Heritage. Two formats are available: Explainer for detailed summaries and Brief for concise overviews. The AI pulls illustrations from content ensuring relevance such as diagrams for technical notes or portraits for biographical sources.
Prompts refine visuals through dropdown menus or text inputs maintaining narrative alignment. Videos export in standard resolutions with audio narration based on the document text. This enhances educational professional and creative workflows by providing illustrated explanations without external design tools. The feature requires a NotebookLM account and processes queries on Google’s servers for accuracy.
What is Nano Banana capable of
Nano Banana built on Gemini 2.5 Flash supports text to image generation image editing multi image composition and conversational refinements. It maintains subject identity across outputs preserving characteristics like facial features and poses in edits. The model handles diverse prompts including outfit changes style transfers and scene blends with consistent quality. Processing occurs with high efficiency generating images in seconds and supporting iterative adjustments.
Advanced features include upscaling to higher resolutions artifact reduction and style application from reference images. The tool integrates SynthID watermarking for origin verification and adheres to Google’s content policies excluding harmful or inappropriate generations. Compatibility spans mobile web and upcoming desktop interfaces with API access for developers through Google AI Studio.
Nano Banana built on Gemini 2.5 Flash supports text to image generation image editing multi image composition and conversational refinements. It maintains subject identity across outputs preserving characteristics like facial features and poses in edits. The model handles diverse prompts including outfit changes style transfers and scene blends with consistent quality. Processing occurs with high efficiency generating images in seconds and supporting iterative adjustments.
Advanced features include upscaling to higher resolutions artifact reduction and style application from reference images. The tool integrates SynthID watermarking for origin verification and adheres to Google’s content policies excluding harmful or inappropriate generations. Compatibility spans mobile web and upcoming desktop interfaces with API access for developers through Google AI Studio.
Nano Banana is available in English for users in the United States and India on Android and iOS devices through the Google app. Access in Search requires opting into AI Mode lab via search settings. Lens integration appears in the app with the Create tab for photo based prompts. NotebookLM access needs a Google account with document uploads limited to 50 sources per overview. Generation limits apply at 50 images per day for free users with higher quotas for Google One subscribers. Outputs download in PNG format with editable metadata. Support languages expand gradually with no timeline specified.
