Google has introduced Gemini Nano Banana 2.1, a new image generation and editing model.
The model is an update to Nano Banana 2 and supports both image generation and conversational image editing. Google says it improves visual quality, prompt following, character consistency, and text rendering across 1K, 2K, and 4K outputs.
One of the main changes is support for multi-image fusion with up to 14 reference images. It can maintain consistency for up to four characters and preserve details from up to 10 objects in a single generation.
Gemini Nano Banana 2.1 also supports wider panoramic image formats, including 1:4, 4:1, 1:8, and 8:1, with fixes for tiling artifacts at 2K and 4K resolutions. Text rendering and infographic layouts have also been improved.
The model can use Google Search and Image Search for grounding. It also offers configurable thinking levels, with minimal, medium, and high options. Medium is the default setting,
Gemini Nano Banana 2.1 is now available through the Gemini app, Google AI Studio, Gemini API, Google Search AI Mode, Google Ads, Google Flow, and Google Stitch.