88Z AI, 8 October — Google DeepMind has released Nano Banana 2.1, an image model built on Gemini 3.6 Flash. The update focuses on improving design completeness, local editing, multi-subject consistency, and visual naturalness, with benchmark scores showing significant gains in infographic design and character consistency.
At a glance
- Nano Banana 2.1 is built on Gemini 3.6 Flash, enhancing design completeness and local editing
- Benchmarks show multi-character consistency rising from 978 to 1106, with an infographic design score of 1048
The update
Google DeepMind has officially launched Nano Banana 2.1, a new model for image generation and conversational editing. Built on the Gemini 3.6 Flash architecture, it carries the model ID gemini-nano-banana-2.1. This release strengthens four core capabilities: improving overall design completeness, enabling more precise mask-based local editing, enhancing consistency among multiple subjects within a single frame, and refining the natural realism of visuals. According to published benchmarks, Nano Banana 2.1 performs strongly on key metrics. Its infographic design score reached 1048, while the multi-character consistency score jumped significantly from 978 to 1106, demonstrating notable progress in handling complex scenes and multi-element compositions.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- 机器之心Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.