Imagenanonano-banana-2

Nano Banana 2

Nano Banana 2 (Gemini 3.1 Flash Image) is Google's most advanced image generation model, combining speed with high-fidelity 4K output and revolutionary character consistency.

Open in Image Lab Browse all models~15 credits per run

Nano Banana guideVerified

Google's flagship image generation. Nano Banana Pro is the current best — strong scene composition, instruction-following, and complex multi-subject layouts.

Strengths

Best instruction-following for image models — Nano Banana actually places objects where you say.
Strong on complex scenes with multiple subjects, relationships, spatial layout.
Edit variant (nano-banana-edit) supports up to 14 reference images.
Reliable for product / commercial work.

Weaknesses

Aesthetic less distinctive than Midjourney (defaults to a Google-style polish).

Best for

Product photography with specific layouts ("bottle on left, ice cubes right")
Multi-subject scenes ("man holding a child, dog at his feet, mountains behind")
Commercial / e-commerce work

Prompting tips

Be EXPLICIT about object placement and relationships.
Numeric counts work ("three apples on the table") better than vague ("some apples").

Parameters

safety_tolerance
string
The safety tolerance level for content moderation. 1 is the most strict (blocks most content), 6 is the least strict.
123456
default: 4
limit_generations
boolean
Experimental parameter to limit the number of generations from each round of prompting to 1. Set to `True` to to disregard any instructions in the prompt regarding the number of images to generate and ignore any intermediate images generated by the model. This may affect gener…
default: true
num_images
int
The number of images to generate.
default: 1
range: 1 … 4
seed
int
The seed for the random number generator.
resolution
string
The resolution of the image to generate.
0.5K1K2K4K
default: 1K
output_format
string
The format of the generated image.
jpegpngwebp
default: png
aspect_ratio
string
The aspect ratio of the generated image. Supports extreme ratios: 4:1, 1:4, 8:1, 1:8. Use "auto" to let the model decide based on the prompt.
auto21:916:93:24:35:41:14:53:42:39:164:11:48:11:8
default: auto
system_prompt
string
Optional system instruction that steers the model's persona and output style across the request. Leave blank to omit; when provided, it is sent as the system instruction to Gemini (or as a system message on OpenAI-compatible providers).
default:
enable_web_search
boolean
Enable web search for the image generation task. This will allow the model to use the latest information from the web to generate the image.
default: false
thinking_level
string
When set, enables model thinking with the given level ('minimal' or 'high') and includes thoughts in the generation. Omit to disable.
minimalhigh

You'll need

A text prompt

Try now

More from nano