Google introduced the Nano Banana AI image generator in August last year, also known as the Gemini 2.5 Flash Image AI model. According to the company, the tool offered faster image creation, improved element-based editing, and stronger character consistency. Later, in November 2025, Google released Nano Banana Pro powered by Gemini 3 Pro, which supported image generation up to 4K resolution. The company described it as an advanced image generation and editing system.
Now, the Mountain View-based tech company has launched Nano Banana 2, which promises Pro-level image quality at faster speeds.
Availability and Access
In a blog post published on Thursday, Google announced the rollout of Nano Banana 2. Users can now access the Gemini 3.1 Flash Image AI model, also called Nano Banana 2, across multiple Google services. In the Gemini app, the new model replaces Nano Banana Pro in the Gemini 3 series Fast, Thinking, and Pro AI modes. However, Google AI Pro and Ultra subscribers can still regenerate images using the older model for specialised tasks.
Nano Banana 2 is now available in 141 countries, including India, and supports eight additional languages. Users can access it through the Google app, mobile browsers, and desktop browsers in Search AI Mode and Google Lens.
Google has also released the model as a preview in AI Studio, the Gemini API, and Vertex AI. Additionally, it has become the default image generation model in Flow, the company’s high-fidelity video creation platform. Flow users can use it without spending any credits. Furthermore, Nano Banana 2 is integrated into Google Ads, where it helps create banners, campaign creatives, and ad suggestions.
Features and Improvements
According to Google, Nano Banana 2 uses Gemini’s advanced knowledge base and pulls real-time information and reference images from web search. As a result, it delivers faster and more accurate subject-focused rendering. It can also create detailed infographics, diagrams, and data visualisations.
The model generates clear and readable text within images, making it suitable for marketing materials, greeting cards, and other creative content. It can also translate and adapt text embedded inside images based on context.
Google claims that Nano Banana 2 quickly produces high-quality photorealistic images. It maintains character consistency for up to five subjects and preserves visual accuracy for up to 14 objects in a single workflow. The system follows instructions more effectively and understands complex prompts with greater precision.
Like Nano Banana Pro, the new model supports image generation up to 4K resolution and delivers brighter lighting effects, richer textures, and sharper details.
