On Thursday, Google rolled out Nano Banana 2, the latest version of its AI image generator, replacing the original Nano Banana across Gemini’s Fast, Thinking and Pro models. The update — also known as Gemini 3.1 Flash Image — is designed to deliver faster image generation while incorporating the advanced world knowledge and reasoning capabilities previously associated with Nano Banana Pro. Google said Nano Banana Pro will remain available for high-fidelity tasks requiring maximum factual accuracy, while Nano Banana 2 will focus on rapid generation, precise instruction following and integrated image-search grounding.
The original Nano Banana launched in August and quickly went viral, followed by Nano Banana Pro in November, built on Gemini 3 Pro. With this new release, Google is merging speed and intelligence into a single model that it says brings “Flash” performance to high-quality visual generation.
According to Google’s blog post, Nano Banana 2 pulls real-time information and images from web search through Gemini, enabling more accurate renderings of specific subjects. The company says the model improves precision text rendering for use cases like marketing mockups or greeting cards and supports translation and localization within images. It also offers enhanced instruction following, subject consistency across up to five characters and 14 objects in a single workflow, and production-ready aspect ratios and resolutions ranging from 512 pixels to 4K.
Google is positioning the update as both a quality and workflow upgrade. The company says Nano Banana 2 improves photorealism with sharper details, richer textures and improved lighting, while also making iterative edits faster. For users who need maximum factual grounding and studio-level control, Nano Banana Pro will remain accessible to Google AI Pro and Ultra subscribers via regeneration options.
The rollout is broad. Nano Banana 2 is available across the Gemini app, Google Search’s AI Mode and Lens, AI Studio and the Gemini API, Vertex AI in Google Cloud, Flow, and Google Ads. In the Gemini app, it replaces Nano Banana Pro as the default image model across core tiers. In Flow, it becomes the default image generation model available to all users at zero credits.
The update lands as AI image and video generation tools continue to gain traction among consumers and creators. OpenAI launched its video generator Sora in 2024, and CEO Sam Altman wrote last year that heavy usage was “melting” the company’s AI processors. Adobe has also pushed AI deeper into its creative tools through Firefly. At the same time, generative media tools have drawn scrutiny over copyright and intellectual property concerns. ByteDance has faced backlash from major Hollywood studios, including Disney and Paramount, over its AI video tool.
Google says it is continuing to expand its provenance efforts alongside the new model. The company is pairing its SynthID watermarking technology with C2PA Content Credentials to help users verify AI-generated content. Since November, Google says its SynthID verification feature in the Gemini app has been used more than 20 million times across languages, and it plans to bring C2PA verification into the app soon.
With Nano Banana 2, Google is refining one of its most popular generative tools rather than introducing an entirely new category. The company is betting that faster generation, improved instruction following and tighter integration across its products will keep momentum going in a rapidly evolving image-generation market.
This analysis is based on reporting from CNBC.
Images courtesy of Google.
This article was generated with AI assistance and reviewed for accuracy and quality.
