‘Gemini 3’-based image creation and editing performance↑
High-quality infographics and diagrams are also provided.
Apply watermark and synth ID… “Enhanced product transparency”
Google’s latest AI image creation and editing tool, ‘Nano Banana Pro’, introduced two days after unveiling its artificial intelligence (AI) model ‘Gemini 3’, is rapidly spreading among global users. Just two days after its launch, actual usage reviews continued to appear in communities such as
In particular, as the ability to simultaneously process old photo restoration, figure production, high-resolution upscaling, and multilingual translation is attracting attention, it is analyzed that changes have begun in the existing generative AI market centered on text generation. However, caution also coexisted as some pointed out errors in transformation and limitations on the free version in the actual work.
According to the information technology (IT) industry on the 23rd, the Nano Banana Pro, which Google released on the 20th (local time), is gaining popularity among users. Nano Banana Pro has improved performance, including text generation and image upscaling up to 4K resolution, based on the new AI model Gemini 3 introduced by Google. Since introducing the existing image creation model Nano Banana last August, users have created more than 5 billion images with Gemini.
Nano Banana Pro goes beyond simple image creation and creates infographics or diagrams based on content and factual information provided by the user. You can also get the educational materials you need to learn new topics. It is also linked to Google search, so the latest information can be reflected in real time. Analysis suggests that this can be useful for commercial image processing as well.
In particular, Google explained that multilingual reasoning capabilities have been improved, making it possible to generate text or localize content in multiple languages. In fact, when “Show me an image of a T-shirt with the ‘Digital Times’ Korean logo on it,” a scene was created showing a man wearing a dark blue T-shirt with the logo engraved on it, standing next to an office building in Korea. At first glance, it looked as natural as an actual photo.
However, errors in Korean were still found in the infographic created with the command (prompt) ‘Draw the process of growing Skin Dapsus as an infographic.’
You can also input up to 14 images and then use them as components to create a new image. In this process, the initially input image is applied to the generated image while maintaining consistency. Advanced creative controls are also available. The ‘Edit by Area’ function has been improved, allowing you to select any part of the image to trim and transform it.
You can also adjust the camera angle, change focus, color correct, and change scene lighting. You can also create scenes that change day into night. It has a variety of screen ratios and supports up to 4K resolution.
Nano Banana Pro is available on the Gemini app. When the free quota is exhausted, it switches to the existing Nano Banana model. Google AI Plus, Pro, and Ultra subscribers can receive higher quotas.
Google explained that products created with generative AI tools are equipped with a digital watermark as standard. This is to strengthen transparency about content sources. In addition, a verification function was added to upload an image to the Gemini app through ‘Synth ID’ and ask whether it was created by Google AI. For example, if you ask, “Was this image made with Google AI?” Gemini checks the Synth ID watermark and uses its own inference capabilities to provide you with the result.
In the future, the scope of support will expand from images to audio and video. However, the watermark is not displayed on the results of professional ‘Ultra’ subscribers and Google AI Studio users.
This is interpreted as a measure to prevent various concerns, such as fake news and fraud, that may arise as generative AI images become more sophisticated. Google said, “In addition to developing our own tools, we are developing source verification technology as a member of the ‘C2PA’ steering committee to strengthen transparency across the industry,” and added, “We will continue to contribute to creating the future of AI transparency.”
Text and photo = Reporter Kim Na-in silkni@dt.co.kr
[저작권자 ⓒ디지털타임스, 무단 전재-재배포, AI 학습 및 활용 금지]
