Nano Banana Pro — Ultra-Fast Lightweight AI

Nano Banana Pro trades speed for significantly enhanced quality and capabilities. Built on Gemini 3 Pro architecture rather than Gemini 2.5 Flash, it offers superior reasoning, dramatically improved text rendering, better character consistency across multiple images, and advanced composition capabilities.

Yes, Nano Banana Pro excels at multilingual text rendering with industry-leading accuracy. The model generates clear, legible text in various languages including English, Chinese, Japanese, Spanish, French, and others with proper character formation and spelling.

Nano Banana Pro can access Google Search to research real-world information before generating images. When you request content requiring current data—like weather infographics, travel guides, recipe visualizations, or educational diagrams—the model searches for accurate information and incorporates it into the visual output.

Nano Banana Pro supports multiple resolutions including 1K (1024×1024), 2K (2048×2048), and 4K (4096×4096) for ultra-high-definition output. Available aspect ratios include square (1:1) for social media, 16:9 for widescreen content, 9:16 for vertical mobile-optimized images, plus specialized ratios like 21:9 for ultra-wide cinematic compositions, 3:2, 4:3, 5:4, 4:5, 3:4, and 2:3.

You can upload up to 14 reference images for a single generation. This extensive multi-image blending capability allows you to combine style guides, character designs, product shots, brand assets, texture references, and other visual elements into cohesive compositions.

Yes, Nano Banana Pro can maintain visual consistency of up to 5 individuals across multiple generations. By uploading reference images of specific characters or people, the model preserves facial features, styling, and appearance across different scenes, poses, and contexts.