Google Launches Nano Banana 2 Image AI Model: Supports 4K Resolution Output, Improves Chinese Character Garbling Issues

robot
Abstract generation in progress

IT House February 27 News, Google has officially launched the Nano Banana2 image model. The new model is built on the Gemini3.1 Flash Image and offers faster response times and stronger understanding capabilities compared to the previous generation, with more accurate interpretation of user prompts.

According to Google, Nano Banana2 shows significant improvements in consistency during generation, and has fixed common issues in earlier versions such as Chinese character garbling, semantic confusion, or strange artifacts appearing in images.

▲ Content generated by other models in Chinese is prone to garbling issues

▲ Nano Banana2 focuses on fixing Chinese character garbling problems

In terms of performance, Nano Banana2’s image output resolution has increased from 2K to a maximum of 4K, capable of handling more complex prompts, supporting up to 5 character identities consistent across multiple images, and blending up to 14 elements within a single image.

In the future, Nano Banana2 will gradually be integrated into Google Gemini, Google Search, Google AI Studio, Google Flow (IT House note: Google’s AI creative tools), paid API services, and Google Ads, replacing the old Nano Banana.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pin