Google has launched the Nano Banana 2, and according to them, they have improved the quality while reducing the overall time taken to generate an image, and the image resolution can be scaled from 512 to 4K resolution for higher production quality. A lot more changes have been made. Let’s check them out.
Image Generation
The Nano Banana 2 creates an image much faster than the Nano Banana Pro in our testing. While the speed is better on the Nano Banana 2, the final output is better on the Nano Banana Pro. While, yeah, the time taken is 3-4s lower, it failed to remove the background and gave me the one with the lines on the back of the image. We had given the same task to ChatGPT Images; it gave us a clearer and better image than the one provided by the Nano Banana 2. The Nano Banana Pro was clearly close to the GPT image, which is weird since the Nano Banana 2 is the successor to the Nano Banana Pro.
Nano Banana 2 is based on the 3.1 Flash, and the Gemini app features a new templates feature. It includes over ten templates, which generate images based on a specific topic. The model utilizes real-world knowledge, which helps to create infographics, turn notes into diagrams, and generate data visualizations. The Nano Banana offers vibrant lightning, richer textures, and sharper details while maintaining the speed of Flash models, allowing us to create excellent images while taking less time to do so. We can maintain character references for up to five characters and 14 objects in a single workflow without altering the appearance of input.
The first image was the one generated from Nano Banana Pro, and the second one was from Nano Banana 2. While, yeah, the second one looks kind of better, the background is worse in the image.
Availability
The Nano Banana 2 is available in the Gemini app, and the new Nano Banana 2 will replace the Nano Banana Pro across the app for the fast, thinking, and pro models. Google AI Pro and Ultra subscribers will keep access to Nano Banana Pro for specialized tasks by regenerating the images. The model is also available in AI mode and Lens through the Google app, as well as mobile and desktop browsers, and includes eight additional languages. The model is available in AI Studio and the Gemini API and also available in Google Antigravity. The model is available in preview in Vertex AI Studio, and the flow is the default image generation model available to all flow users.
Security
The Nano Banana 2 has enhanced security features. When I tried the image on the Nano Banana 2 and tried to modify an image of mine with some changes for the dress, etc., it said that it couldn’t modify the image, which is a great addition for all the people, since now AI can’t modify any details of your personal image and can’t use it for data training. The Nano Banana Pro allowed for this modification; it’s a great addition overall. The tools that are used to identify AI are evolving. Now the SynthID and C2PA context credentials are combined to determine if AI was used, but how?
Speed
The speed has been improved overall. We tried both models with various prompts. We found that the new Nano Banana 2 was significantly faster in some cases. However, in some instances, the speed was faster; it was about 5% faster, and the image quality was subpar. In our previous testing, the Nano Banana Pro in the AI Studio produced excellent quality images, but after reaching a certain token limit, its performance would decline. It would take 70s to generate an image, and we generated around 10 images. We see that the performance is awesome; it still takes roughly 20-25 seconds to generate an image, which is excellent. So, the context performance has improved.
We hope that you have enjoyed this article. For more information on electric vehicles, artificial intelligence, innovation, and the technology that will shape the future, continue to browse TechOrbis. At TechOrbis, our aim is to provide insightful analysis, honest views, and meaningful perspectives that extend beyond the news.
Follow TechOrbis on X/Twitter, Instagram, LinkedIn, and Telegram to keep up with the latest news, trends, and tech talk. As a growing and independent source of news, every view, share, and mention helps us reach more people and continue to provide meaningful tech journalism.
