Google has unveiled Nano Banana 2, its latest AI image generation model. The new model, also known as Gemini 3.1 Flash Image, builds on the success of the original Nano Banana, which went viral for its image editing skills, and Nano Banana Pro, which offered advanced intelligence and studio-quality creative control. According to the company, Nano Banana 2 brings the ‘best of both worlds to users across Google.’
Nano Banana 2 is designed for faster image generation and editing, making quick edits and iteration possible. It uses Gemini’s world knowledge and real-time web information to create more accurate visuals of specific subjects. This added intelligence also improves its ability to generate infographics, simple diagrams and data-based visuals, according to Google.
Also read: Google redesigns Flow, combines image and video creation in one unified workspace
The model is also said to be better at handling text within images. This means that users can create clear, readable text for marketing materials, greeting cards and social media posts. It can also translate and localise text inside images.
Google has also improved creative control in Nano Banana 2. The model can maintain consistent appearances for up to five characters and keep details accurate for up to 14 objects in a single workflow. The company also claims that Nano Banana 2 follows detailed instructions more closely, ensuring the final image matches what the user describes. It also supports different aspect ratios and resolutions from 512 pixels up to 4K.
Nano Banana 2 is rolling out across Google products. It is available in the Gemini app, through Google Search in AI Mode and Lens, in AI Studio and the Gemini API, in Vertex AI on Google Cloud, in Flow, and inside Google Ads. Users can start trying the new model today through these platforms.