Text-to-image · all categories · (Gemini 3.1 Flash Image)
95% CI 1,129.99–1,149.99 · n=16,075
Nano Banana 2 makes high quality image generation and editing mainstream. The model also introduces Grounding with Google Image Search to enable long-tail entity recognition and enhance visual understanding.
Text-to-image · all categories · (Gemini 3.1 Flash Image)
95% CI 1,129.99–1,149.99 · n=16,075
512px
Web search adds $0.015; high thinking adds $0.002.
1K
Web search adds $0.015; high thinking adds $0.002.
2K
Web search adds $0.015; high thinking adds $0.002.
4K
Web search adds $0.015; high thinking adds $0.002.
512 output image
Each input image adds $0.0672.
1K output image
Each input image adds $0.0672.
2K output image
Each input image adds $0.0672.
4K output image
Each input image adds $0.0672.
google-vertex/us · output image tokens · variant=default
Conditional rateBilling component; not a complete generation price.
google-ai-studio · output image tokens · variant=default
Conditional rateBilling component; not a complete generation price.
google-vertex/global · output image tokens · variant=default
Conditional rateBilling component; not a complete generation price.
512px
Web search adds $0.015; high thinking adds $0.002.
1K
Web search adds $0.015; high thinking adds $0.002.
2K
Web search adds $0.015; high thinking adds $0.002.
4K
Web search adds $0.015; high thinking adds $0.002.
google/gemini-3.1-flash-imagefal-ai/nano-banana-2google/nano-banana-2fal-ai/nano-banana-2gemini-3.1-flash-image3180162e-693d-487d-adbb-721f859f768dUpdated Oct 21, 2026
Edits at up to 4K using image, video or PDF context; supports extra-wide 8:1 and extra-tall 1:8 outputs.
Nano Banana 2.1 (Gemini Nano Banana 2.1) is Google's image generation and editing model on the Flash tier, succeeding Nano Banana 2 and Nano Banana Pro. It improves product recontextualization,...
Generate realistic virtual try-on images from a person image and a clothing product image.
Nano banana lite is the efficiency-focused model in the image generation family. Sub-2 second latency with cost-effective generation and editing, fast multi-turn local edits, and 14 supported aspect ratios.