Google launches Nano Banana 2 with faster, high-quality image generation

Google has launched the next generation of its image generation model, the Nano Banana 2, with the ability to generate high-quality images at “lightning-fast speed.” The new image model is powered by the Gemini 3.1 Flash that claims to generate more realistic images in comparison to its predecessor. 

Bringing the lightweight Gemini Flash AI model into its image generator will make the process quicker and more responsive for back-and-forth iteration. Google claims that instead of relying on training data, the model can capture real-time information from the web, like a specific landmark, person, or product, to generate more accurate images.

Google said, “This deep understanding also helps you create infographics, turn notes into diagrams and generate data visualisations.” Furthermore, the new image model also claims to generate accurate text for marketing mockups or greeting cards. It will also have the ability to translate and localise text.

Nano Banana 2: What’s new and upgraded

According to the Google blog post, the Nano Banana 2 can now maintain the appearance of up to five characters and up to 14 objects within a single workflow. Google also brings greater control over choosing aspect ratios and resolutions, ranging from 512px all the way up to 4K.

The company also claims to bring a visual fidelity upgrade with better lighting, richer textures, and sharper details, all with faster generation capabilities. In addition, the model can also manage strict to complex prompts, allowing users to ask for finer details and specific nuances.

Google says that Nano Banana 2 is built for “ rapid generation, precise instruction following and integrated image-search.”

Nano Banana 2 availability 

The Nano Banana 2 has started to roll out across Google products, including the Gemini app, Search (AI Mode and Lens), AI Studio, Gemini API, Google Cloud, Flow, and Google Ads.