How to use Nano Banana on Runway
Add your prompt and references
Describe the image you want in full sentences — Nano Banana is built on Gemini's language understanding, so narrative prompts outperform keyword lists. Add up to 3 reference images to guide the output.
Pick your aspect ratio
Choose from 21:9, 16:9, 4:3, 1:1, 3:4 or 9:16 to match wherever the image will live, from cinematic banners to vertical social posts.
Generate and iterate
Generate fast 1K outputs, then refine with follow-up edits — swap backgrounds, restyle a subject or composite elements from your references.
What you can do with Nano Banana
Generate at speed
Optimized for fast, high-volume image generation — the right choice when you need lots of on-brief images quickly.
Edit with words
Change backgrounds, restyle subjects or adjust details in an existing image with a plain-language instruction.
Composite from references
Combine up to 3 reference images into one output — keep a product, face or logo consistent while changing everything around it.
Generate your first image with Nano Banana on Runway.

Runway Dev
Build with the Nano Banana API
Ship Nano Banana inside your own product through Runway Dev, alongside the best models from Runway and other labs. The text to image API endpoint generates and edits images.
Use the APIMore models on Runway
Runway gives you access to the world's leading video, image and audio models — Runway's own frontier models plus the best from Google, OpenAI, Kling, ByteDance and more — all in one workspace.
Nano Banana Pro
Production-grade image generation and editing at up to 4K.
Explore the modelNano Banana Pro 2
Google's latest image model — the best balance of quality, speed and intelligence.
Explore the modelGPT Image 2
OpenAI's image model with exceptional in-image text rendering.
Explore the modelSeedance 2.0
Multi-modal video generation with sound from text, images, video or audio.
Explore the modelTry Nano Banana on Runway today.
Generate, edit and composite images in seconds — alongside every other leading AI model, in one workspace.
