
Nano Banana 2
Best for group photos, multi-object scenes, and editing one person or object. People and objects stay distinct; change a detail without redrawing the whole frame.
- Text to Image
- Image to Image
AI Model Catalog
Compare six AI image models and two AI video models in one workspace. Every model shares one credit wallet, with 10 free daily credits to get started.

Best for group photos, multi-object scenes, and editing one person or object. People and objects stay distinct; change a detail without redrawing the whole frame.

Best for posters, infographics, storyboards, and UI with small type. CJK text stays clear; ultra-wide to ultra-tall frames come out ready to use.

Best for product shots, hero-art extensions, and drafts you keep refining. Blend up to three reference images in a single pass.

Best for poster typos, product copy swaps, and character variants. One prompt can change style or background while keeping the original type.

Best for photoreal shots, Chinese street scenes, and storefront signage. Perspective stays solid, and Chinese shop signs stay readable.

Best for character portraits, anime storyboards, and illustration. Faces and outfits stay on-model; hands and feet hold up better in action poses.

Choose 2.0 or 2.5 with 5, 10, or 15 second generation and up to 1080P output; 2.5 also reaches 20, 25, or 30 seconds.

Best for ads, brand spots, and motion posters with sound. Assign camera, character, and vocals separately; logos and captions hold better.