Google's image-generating AI solves a tough problem

Google has just launched Nano Banana Pro, its latest photo creation and editing model. This is the next AI model introduced by the company this week after Gemini 3. Nano Banana Pro is built on Gemini 3 Pro. Google emphasizes that this is the “most advanced photo editing model” the company has ever released. The model’s improvements revolve around photo editing, creating legible letters and maintaining consistency in design.

Users can experience Nano Banana Pro in the Gemini app (select the Create Image > Brainstorm function). Free accounts can use the model with a limited number of times, then switch back to the old version.

Note-taking app NotebookLM will also integrate Nano Banana Pro, alongside Google's developer, enterprise, and content creator products.

In the US, AI Pro and AI Ultra users can experience Nano Banana Pro in Google Search AI mode. Additionally, the new model is also integrated into a number of popular design apps such as Canva, Figma, and Photoshop.

Google says that with the Gemini 3's enhanced reasoning capabilities, Nano Banana Pro supports creating illustrations with visual content and easier-to-read fonts. For example, this technology can turn recipe text into illustrated charts, or visualize real-time information such as weather or sports .

Text generation is one of the improvements that Google emphasizes on Nano Banana Pro. Compared to the previous version, this model can create accurate, easy-to-read text, supporting translation into many languages. Users can even create text with special effects and layouts, suitable for use in article design, advertising posters.

Nano Banana Pro also improves design consistency. According to Google, users can provide up to 14 different images and have the model arrange them as desired, while preserving the features of up to five images. The tool retains the look and feel of friends, family, and pets regardless of hairstyles or outfits.

Users can request to edit details in existing photos, change shooting angle, ratio, color adjustment, focus position and even lighting, as if taking a photo directly with the camera.

Google emphasizes that images created by Nano Banana are tagged with a hidden SynthID to identify AI content. In the new version, Gemini allows uploading images and asks if the image was generated by Google AI. The company plans to expand the feature to audio and video in the future. In addition to SynthID, images created by Google's model will still have a watermark when using the free plan or Google AI Pro. The watermark will be removed for AI Ultra subscribers.

Since its debut in August, Nano Banana has garnered attention for its ability to create highly realistic images of pets, people, and landscapes. In September, Josh Woodward, VP of Google Labs and Gemini, said Nano Banana helped the Gemini app attract 13 million new users in four days. Woodward said Nano Banana Pro is capable of performing better than the original version, highlighting features like creating infographics, presentation slides, and ensuring consistency between images.

Google is looking to gain an advantage over OpenAI in the generative AI race, which has exploded since ChatGPT launched in 2022. Last week, OpenAI announced version GPT-5.1, which focuses on improving user interaction. ChatGPT is currently the top free app on the US App Store, while Gemini is No. 2.

When Gemini 3 launched, Google announced that the Gemini app had more than 650 million monthly users, and AI Overviews had 2 billion monthly users. In October, OpenAI CEO Sam Altman said ChatGPT had reached 800 million weekly users. Many people subscribe to Gemini plans to increase their limits on advanced model usage.