2025-11-21 02:10:00+08
Google has unveiled its next-generation image generation model, Nano Banana Pro, built on the newly announced Gemini 3 architecture. The model supports up to 4K resolution, seamless blending of up to 14 objects, and maintains identity consistency across up to 5 people in a single image. It also introduces real-time web search integration and dual watermarking via SynthID and C2PA content credentials.
Pricing has been significantly increased:
Professional-grade control: Users can specify camera angles, depth of field, focus points, lighting effects, and color grading directly in prompts. By default, the system delivers six high-fidelity outputs for selection.
Web-connected generation: The model can perform real-time web searches to fetch recipes, news, stock prices, or other live data and automatically generate corresponding infographics. This capability is already integrated into NotebookLM, Flow, Google Slides, and Vids.
Dual watermarking: Every generated image embeds both an invisible SynthID watermark and a C2PA-compliant content credential. Users can upload any image to the Gemini App to verify whether it was generated or modified by a Google AI model.
Developers can integrate Nano Banana Pro via the Gemini API, Google AI Studio, and Workspace plugins, with billing based on image count and resolution.