# Nano Banana Pro is Here - New Features Unlocked! Source: https://www.youtube.com/watch?v=2jkpEGk6hDE Recap page: https://rapidrecap.app/video/2jkpEGk6hDE Generated: 2025-11-20T16:14:52.774+00:00 --- ## Quick Overview Google DeepMind officially launched Nano Banana Pro, an upgraded AI image model that offers superior image generation and editing capabilities compared to its predecessor, Nano Banana, by providing better consistency, more control, and advanced features like style transfer and multi-image blending, though some advanced features like style transfer or specific aspect ratios are reserved for the Ultra subscription tier. **Key Points:** - Google DeepMind officially released Nano Banana Pro, the successor to Nano Banana, on November 20th, featuring enhanced AI image generation and editing. - Nano Banana Pro offers better consistency, more control over prompting, and eliminates the visible watermarks found on the previous model's outputs for Ultra users. - The model excels at complex editing tasks, such as style transfer (applying the watercolor style of one image to another while preserving facial details) and multi-image blending (combining up to six images into one cohesive poster). - Advanced features like style transfer and the ability to precisely control aspect ratios consistently across multiple images (1:1, 16:9, 9:16, 3:2) are available through the paid Gemini Advanced tiers (Pro and Ultra). - The video production tool, Flow, also received an update to Veo 3.1, which supports Nano Banana Pro features, allowing users to generate cinematic stories and product mockups. - The demonstration showed Nano Banana Pro accurately executing complex requests, such as turning a group photo into a futuristic magazine cover and turning a sketch into a photorealistic 3D render of a gadget. ![Screenshot at 00:58: The screen displays the core functionality of Nano Banana Pro, showing an input image of fruit being successfully transformed into an output image showing only bananas, demonstrating its editing capability.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-00-58.png) **Context:** The video provides an overview and demonstration of new features launched by Google DeepMind for its AI image generation models, specifically focusing on the upgrade from Nano Banana to Nano Banana Pro. The presenter, who had early access, walks through various capabilities, including style transfer, multi-image blending, and improved consistency across different aspect ratios, often comparing the new Pro version against the previous iteration and showing how it integrates with other Google AI tools like Flow and Gemini Advanced. ## Detailed Analysis Google DeepMind officially launched Nano Banana Pro on November 20th, succeeding Nano Banana, and offering significant improvements in image generation and editing. The presenter highlights that the new model provides better consistency, greater control through doodling instructions, and superior realism, especially when blending multiple images or applying style transfer. For instance, it successfully applied the watercolor style of a landscape painting to a group photo while keeping facial details recognizable, and it accurately rendered a complex prompt combining five disparate images (dog, surfboard, city skyline, desert road, neon signs) into a cohesive sci-fi movie poster. The ability to generate consistent aspect ratios (1:1, 16:9, 9:16, 3:2) from the same base image was also demonstrated successfully. The presenter notes that while the basic Nano Banana model is free, the most advanced features, like style transfer and consistent aspect ratio generation, are accessible through the paid Gemini Advanced tiers (Pro and Ultra). Furthermore, the update extends to Google's video generation tool, Flow, which now supports Veo 3.1 and Nano Banana Pro, allowing for high-quality, consistent video creation from prompts and existing assets. ### Nano Banana Pro Key Upgrades - Eliminates visible watermarks for Ultra users - Offers better consistency and control - Enables multi-image blending up to six sources - Advanced style transfer demonstrated successfully. ### Image Editing Capabilities - Successfully applied watercolor style to a portrait while maintaining facial features - Accurately followed complex multi-image blending prompts (dog, surfboard, city skyline, etc.) ### Video Generation Integration - Nano Banana Pro features are integrated into Google's Flow tool (via Veo 3.1) for generating cinematic stories and mockups. ### Advanced Prompting & Control - Successfully generated complex textual logo designs (e.g., 'WOLF' in multiple languages/materials) and product renders from sketches (e.g., the AEGIS wrist monitor). ### Aspect Ratio Consistency - Successfully generated the same landscape scene in 1:1, 16:9, 9:16, and 3:2 ratios while maintaining composition. ### Subscription Tiers - Free tier offers basic access; Pro and Ultra tiers unlock advanced features like style transfer and deeper access to models. ![Screenshot at 00:01: Demonstration of Nano Banana replacing all fruits in a bowl with bananas, showcasing editing capability.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-00-01.png) ![Screenshot at 00:14: Visual introduction of the 'Nano Banana Pro' model, personified by a figure with the Google logo as a head holding a banana.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-00-14.png) ![Screenshot at 00:50: The Google DeepMind logo, indicating the source of the AI technology being discussed.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-00-50.png) ![Screenshot at 01:11: Demonstration of style transfer, applying a retro '80s aesthetic to a photo of a woman.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-01-11.png) ![Screenshot at 01:35: Demonstration of the model's ability to generate images based on complex text prompts, showing the 'Ignition Sequence' output.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-01-35.png) ![Screenshot at 03:07: Comparison showing the difference between watermarked and non-watermarked outputs to illustrate the removal of SynthID watermarking for some users.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-03-07.png) ![Screenshot at 04:47: Demonstration of image editing via sketching, where the user doodles bunny ears and eyes onto the Apple Vision Pro image.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-04-47.png) ![Screenshot at 08:08: Demonstration of image analysis, where Gemini correctly annotates landmarks around Petco Park in San Diego.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-08-08.png) ![Screenshot at 09:59: Demonstration of multimodal input where the AI successfully combines five disparate images into one cohesive 'sci-fi road trip movie' poster.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-09-59.png) ![Screenshot at 13:15: A visual comparison showing the presenter giving a thumbs-up next to an image of Mr. Burns, suggesting approval of a feature.](https://ss.rapidrecap.app/screens/2jkpEGk6hDE/00-13-15.png)