Nano Banana Pro

nano-banana-pro

Image generationby Google

Nano Banana Pro (or Gemini 3 Pro Image) is Google's new state-of-the-art image generation and editing model. Use Nano Banana Pro with an API.

images

Try in Studio →

Details

Accepts: text + image

Pricing

Input: 250 cr / 1M tokens
Output: 1500 cr / 1M tokens

Prices in credits (1 credit = $0.01).

Data schema

approximate

Input

Field	Type	Description
n	integer	How many images to generate.
mask	string	Optional mask for the edit; transparent pixels mark the area to change.
image	string	Image to edit — a URL, a data URI, or an artifact reference. Supplying one turns the call into an edit.
promptrequired	string	The text prompt. Describes the image to generate, or the change to make when an image is supplied.
image_size	string	Requested output size.
aspect_ratio	string	Aspect ratio of the result, e.g. 16:9.
output_format	string	Encoding of the returned image.
resolution	—enum: 1K, 2K, 4K	Resolution of the generated image
image_input	array	Input images to transform or use as reference (supports up to 14 images)
safety_filter_level	—enum: block_low_and_above, block_medium_and_above, block_only_high	block_low_and_above is strictest, block_medium_and_above blocks some prompts, block_only_high is most permissive but some prompts will still be blocked
allow_fallback_model	boolean	Fallback to another model (currently bytedance/seedream-5) if Nano Banana Pro is at capacity.
seed	—	The seed for the random number generator.
top_p	—	Nucleus sampling probability. Omit to use the provider default of 0.95. Explicit values use a direct Google or Vertex API.
context	array	Ordered text, original images, PDF or UTF-8 text documents appended to the current prompt. Uses a native provider. Images count toward the total 14-reference limit.
history	array	Previous native user/model turns. Replay a returned conversation's messages unchanged, including thought signatures; the current prompt is appended as a new user turn.
sync_mode	boolean	If `True`, the media will be returned as a data URI and the output data won't be available in the request history.
image_urls	array	The URLs of the images to use for image-to-image generation or image editing.
num_images	integer	The number of images to generate.
temperature	—	Sampling temperature. Omit to use the provider default of 1. Explicit values use a direct Google or Vertex API.
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).
return_history	boolean	Return reusable native conversations with original image parts and provider thought signatures. Uses a native provider. Each generated conversation is independent; send its messages as history for the next turn.
thinking_level	—	Thinking effort. Pro Image supports only high; omit to retain its provider default. An explicit level requests summaries unless include_thoughts overrides it. Thinking cannot be disabled.
include_thoughts	—	Whether to include thought summaries. Omit to preserve the existing behavior: an explicit thinking_level requests summaries. Set limit_generations=false to return summary text in description. This does not disable thinking or return intermediate thought images.
safety_tolerance	stringenum: 1, 2, 3, 4, 5, 6	The safety tolerance level for content moderation. 1 is the most strict (blocks most content), 6 is the least strict.
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.
image_passthrough	boolean	Send reference images and return provider images without resizing or re-encoding, preserving transparency and metadata. Uses direct Google or Vertex APIs; the returned image format may differ from output_format.
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.
max_output_tokens	—	Maximum output tokens for the generation. Omit to retain the provider default. Explicit values use a direct Google or Vertex API.

Output

Field	Type	Description
images	array	Generated images ([{ url }]).