Google Launches Nano Banana 2, Now Top-Ranked AI Image Generation Model
They delivered a hit and now want to follow up quickly: Google has unveiled its latest image generation model Nano Banana 2 (technically it’s Gemini 3.1 Flash Image) today. The model is intended to combine the advanced features of the already available Pro version with the speed of the Flash variant and will become the standard in various Google products.
According to the Arena.ai ranking, Nano Banana 2 is currently among the best, and possibly even the best AI model in terms of image generation:
Development of the Nano Banana Series
The Nano Banana series started in August 2025 with the first model, which according to Google went viral and prompted millions of users to generate images, particularly in countries like India. In November, Nano Banana Pro followed with advanced features for professional applications. Now Google is presenting Nano Banana 2 as a middle-ground solution.
Announced Key Features
Google claims that Nano Banana 2 brings the following core capabilities:
Advanced Knowledge Base and Text Processing
- Global Knowledge: The model accesses Gemini’s knowledge database and uses real-time information from web search for more accurate representations
- Infographics and Diagrams: Creation of data visualizations and conversion of notes into diagrams
- Text Generation and Translation: Precise, readable text rendering in images as well as translation and localization of text within images
Creative Control and Image Quality
- Character Consistency: Maintaining similarity of up to five characters and 14 objects in a workflow
- Precise Instruction Following: Improved implementation of complex requests with specific nuances
- Production-Ready Specifications: Resolutions from 512px to 4K in various aspect ratios
- Visual Quality: Vibrant lighting, richer textures, and sharper details
Availability in Google Products
According to the announcement, Nano Banana 2 will be rolled out in the following Google services:
- Gemini App: Replaces Nano Banana Pro in Fast, Thinking, and Pro modes. Subscribers to Google AI Pro and Ultra can continue to access Nano Banana Pro
- Google Search: In AI mode and in Lens via the Google app as well as mobile and desktop browsers in 141 countries and eight additional languages
- AI Studio and API: Available as a preview version
- Google Cloud: Preview via Gemini API in Vertex AI
- Flow: New standard image generation model for all Flow users
- Ads: Support for campaign creation
Origin Verification and Authentication
Google is committed to transparency in AI-generated content. The company combines its SynthID technology with C2PA Content Credentials to show users not only whether AI was used, but also how.
According to Google, the SynthID verification feature introduced in November has already been used over 20 million times in the Gemini app. C2PA verification is expected to be integrated into the Gemini app soon as well.
Positioning in the Product Portfolio
With the introduction of Nano Banana 2, Google now offers two main options: Nano Banana Pro for high-resolution tasks with maximum factual accuracy and Nano Banana 2 for fast generation with precise instruction following and integrated image search foundation.
The announcement comes approximately three months after the release of Nano Banana Pro and nine months after the original Nano Banana model.
