Google has released Nano Banana 2.1, an updated image generation and editing model built on the Gemini 3.6 Image architecture. The new version focuses on enhancing precision when modifying specific image segments, allowing users to alter selected areas without regenerating the entire scene.
A major improvement in this release is subject consistency. The model maintains the appearance of people and objects more effectively during iterative edits or when creating sequences of related images. Additionally, developers have reported enhancements in overall visual composition and output quality.
The model retains core features from its predecessor, including improved text rendering and the integration of real-world knowledge. It supports multimodal inputs, processing text, images, audio, and video data to generate outputs in 1K, 2K, and 4K resolutions.
Nano Banana 2.1 is currently available across several platforms, including the Gemini app, the AI Mode in Google Search, Google AI Studio, Flow, Stitch, Google Ads, and the Gemini Enterprise Platform. For developers, the model is accessible via the identifier gemini-nano-banana-2.1.
