
Google's premium AI image generation model.
94% text rendering accuracy, advanced "Thinking" mode, and up to 5 character references for unmatched creative control.
The most capable AI image model with reasoning-powered generation

Nano Banana Pro (Gemini 3 Pro Image) is Google's premium AI image generation model. It features advanced "Thinking" reasoning that plans compositions before generating, achieving 94% text rendering accuracy and supporting up to 5 character references in a single workflow.
Industry-leading text accuracy in generated images. Perfect for marketing materials, posters, and branded content with precise typography.
Pro model reasons about composition, lighting, and spatial relationships before generating — producing more coherent and intentional results.
Maintain consistency across up to 5 characters simultaneously. Ideal for storyboarding, comics, and multi-character scenes.
Generate images at 1K or 2K resolution with exceptional detail, rich textures, and professional-grade output quality.
An advanced reasoning step that plans spatial layout, lighting, and subject relationships before pixel generation — resulting in more coherent compositions.
Near-perfect text rendering within images across multiple languages, making it the most reliable model for text-in-image generation.
Track up to 5 distinct characters with consistent appearance across multiple scenes — perfect for visual storytelling and brand mascots.
Accept multiple reference images as input for style matching, character consistency, and compositional guidance.
Advanced color grading and lighting simulation for cinema-quality output that matches professional photography standards.
Reasoning-powered spatial understanding handles multi-subject scenes, proper occlusion, and natural depth-of-field effects.
Experience the most capable AI image model with Thinking reasoning and 94% text accuracy.