Today, Google introduced the latest iteration of its acclaimed image generation model, Nano Banana 2, also known as Gemini 3.1 Flash Image. This upgraded model is designed to produce more lifelike images than its predecessor and will serve as the default option in the Gemini app across its Fast, Thinking, and Pro modes.
Since its initial launch in August 2025, the original Nano Banana model has inspired users to create millions of images within the Gemini app, particularly in vibrant markets like India. The subsequent release of Nano Banana Pro in November 2025 further enhanced user capabilities, allowing for the creation of highly detailed images.
The newly launched Nano Banana 2 combines the high-fidelity features of the Pro model with significantly faster image generation. Users can now create images with resolutions ranging from 512px to 4K, accommodating various aspect ratios.
One of the standout features of Nano Banana 2 is its ability to maintain character consistency for up to five characters while ensuring fidelity for as many as 14 objects in a single workflow. This enhancement allows for richer storytelling through nuanced image generation. Additionally, users can craft images with more vibrant lighting, intricate textures, and sharper details.
With this launch, Nano Banana 2 will become the standard model for image generation in all applications within the Gemini ecosystem, including the video editing tool, Flow. Furthermore, it will be integrated into Google Search results via Google Lens and AI Mode, enhancing user experience across 141 countries on both desktop and mobile platforms.
For those subscribed to Google's premium AI plans, such as AI Pro and Ultra, the option to utilize Nano Banana Pro remains available for specialized tasks, allowing for image regeneration through a simple menu option.
Developers can access Nano Banana 2 in preview mode via the Gemini API, Gemini CLI, and Vertex API, as well as through AI Studio and the recently released development tool, Antigravity. All images produced with this model will feature a SynthID watermark, a mark indicating AI-generated content. Additionally, these images will be compatible with C2PA Content Credentials, a collaborative initiative among major tech companies like Adobe, Microsoft, and Meta. Since the introduction of SynthID verification in the Gemini app, it has been utilized over 20 million times, underscoring its growing importance in the digital landscape.