Google launches Nano Banana 2: Explore key features and how to use it in Gemini, Lens, Flow

Google on Thursday launched Nano Banana 2, the successor to its viral AI image generation model, promising faster performance and improved features across its suite of AI tools. The model is being integrated into Google products including the Gemini app, AI Mode, Lens in Search, and Flow, its AI-powered video tool.

The rollout marks Google’s continued push to lead the AI sector, strengthening its position against competitors such as OpenAI’s ChatGPT. The company’s AI advancements have coincided with a 47% surge in its stock over the past six months.

Viral success of Nano Banana models

The original Nano Banana AI image editor was launched in August and quickly became a viral sensation. It drew 13 million first-time users to the Gemini app in just four days in September and had generated over 5 billion images by mid-October. Google followed this with the release of Nano Banana Pro in November. Nano Banana 2 builds on this evolution by combining the best aspects of its predecessors into a faster, more capable model.

The evolution: meet the family

• Nano Banana (Aug 2025): The original viral sensation, fast and fun but struggled with complex text, fingers, and spatial reasoning.
  
  
• Nano Banana Pro (Nov 2025): The "heavy lifter," offering 4K quality and 16-bit color, though slower and more expensive to run.
  
  
• Nano Banana 2 (Feb 26): Combines Pro features like 4K resolution and accurate text with rapid image generation in seconds.
  
  

Key features of Nano Banana 2

• Subject consistency: Maintain up to five characters across different scenes, ideal for storytelling and comics.
  
  
• Near-perfect text rendering: Handles complex labels, signs, and full infographics with typographic precision.
  
  
• Real-time web grounding: References real-world locations, events, and objects using Google Search for accuracy.
  
  
• Production-ready specs: Supports 4K resolution and ten aspect ratios, from TikTok verticals to cinematic widescreens.
  
  
• Semantic editing: Modify images with natural language, e.g., “Change the red car to a blue SUV and make it sunset.”
  
  

Getting the most out of Nano Banana 2

Users are encouraged to treat the model like a creative director rather than a search engine. Techniques include:

• Plan → Evaluate → Improve: Give multi-step instructions to build complex scenes.
  
  
• Iterate via conversation: Refine lighting, objects, and camera angles through follow-ups.
  
  
• Leverage technical knowledge: Use photography terms like “Rembrandt lighting” or “16-bit color depth” for high-fidelity results.
  
  
• Character sheets: Create multi-angle references to keep subjects consistent across scenes.
  
  

Applications and availability

Nano Banana 2 is available today across the Gemini app, Google Lens, and Vertex AI. It has also been integrated into Flow, Google’s AI-powered video tool, to maintain consistency between generated video frames and still images.

AI engagement and competitive impact

In November, Google released Gemini 3, which significantly boosted user engagement and prompted rival OpenAI to issue an internal “code red” to accelerate development. By the end of December, the Gemini app had over 750 million monthly active users, reflecting strong adoption of Google’s AI offerings.