nano-banana:official corresponds to Google Gemini 2.5 Flash Image, suitable for generating images from text concepts and using reference images for scene changes, color adjustments, and asset blending. Its practical focus is on continuously creating visual variations around the same product or character, enabling a cohesive workflow for creative validation, product presentation, and marketing asset revisions without having to rethink the entire image each time.
Submit requests to the public API at api.acedata.cloud using the documented parameters, then use the results in your application.
Input parameters and result formats vary by service. Use the public API for this model and follow its guide for generation, task retrieval and editing operations.
The model positioning describes its image creation capabilities; aspect ratio, quantity, and task processing options describe the scope of this platform's API endpoint and do not represent native pixel specifications.
Use a product photo or character image to establish a visual baseline, then describe new scenes, lighting, and compositions, making it suitable for creating a continuous series of assets around the same subject. Prompts should clearly specify the shape, color scheme, and identifying features to retain, so changes focus on the background and atmosphere rather than rewriting both the subject and scene at once.
Editing does not need to restart from a description of the entire image. After providing the original image, you can request a background replacement, lighting adjustments, object color changes, or the addition of new elements. Express preserved areas separately from modification goals, and iterate gradually according to local needs to make it easier to assess whether each change aligns with the creative direction.
Using product, background, or style references, the model can attempt to organize different assets into a unified image. The key is to explain the role of each image: which determines the subject, which provides the environment, and which only informs the color palette. This prevents multiple assets from competing for compositional focus and is suitable for brand visuals and product scene exploration.
Input clear product photos and describe target scenes such as wooden tables, bathrooms, or office environments to generate contextual images suitable for product display. You can also experiment with different backgrounds and colorways for the same product, delivering a set of visual options for operations teams to review; check packaging text, outlines, and key structures before formal use.
Provide an event theme, brand references, and an accurate short title; establish the key visual first, then create landscape or portrait assets separately. Suitable for exploring holiday themes, advertising compositions, and cover directions. Keep text brief, specify title hierarchy and whitespace placement, and check glyphs and layout after generation.
Use character reference images together with storyboard descriptions to generate images in different scenes, poses, or shots for storyboards, event mascots, and content planning. Reuse the same identity feature descriptions each time, use selected images as subsequent references, and deliver continuous visual drafts for discussion.
If the task focuses on product scene changes, visual variations, and creative drafts, the fast generation and editing positioning of nano-banana:official is better suited to this workflow. Use it first to establish the subject, composition, and atmosphere, then decide whether to proceed to detailed production. Do not interpret fast positioning as a fixed generation time in seconds, and do not skip checking product details and text within images.
Nano Banana refers to Gemini 2.5 Flash Image and is not equivalent to Nano Banana Pro or Nano Banana 2. If requirements emphasize high-precision final output or specific high-resolution delivery, compare the detailed specifications of the corresponding models; the :official suffix in this entry is used to select the invocation ID, does not indicate switching to Pro, and does not automatically add next-generation capabilities.
Choose generate for text-based creation; choose edit to modify existing assets, provide reference images with image_urls, and separately describe what to preserve and what to change.
Specify model=nano-banana:official, action, and prompt for /nano-banana/images; start with aspect_ratio=1:1, resolution=1K, and count=1, setting the aspect ratio and resolution separately.
Retrieve images from data[].image_url; for asynchronous requests, query with or receive callbacks via task_id. When continuing edits, pass the selected image again and narrow the scope of changes in each round.
Preserve the reference image's face shape, hairstyle, clothing, and composition; change the daytime lighting to warm evening lighting, and change the background to a quiet study.
Compare how well the face and clothing are preserved, then check the shadows and background; fully enter the call ID with its suffix to facilitate alignment of task and cost records.
It is the specific ID used to call Nano Banana; the corresponding public model name is Gemini 2.5 Flash Image. Nano Banana is the model's nickname; :official does not constitute a new Google model name, nor is it equivalent to Nano Banana Pro. Use the full ID when integrating.
Yes. Select generate and provide a prompt describing the subject, environment, lighting, composition, and style to start text-to-image generation. If you want to match an existing product or brand visual, you can also add image references; choose the aspect ratio based on the actual use case, such as a cover, banner, or vertical content.
Use edit, provide the images to process through image_urls, and describe the relationship between the assets in the prompt. For example, retain the product's shape and replace the environment with a morning-lit wooden table scene. When combining multiple images, clearly specify the purpose of the subject image and background image rather than writing only vague compositing requirements.
When creating consecutive variants, it is recommended to use the image selected in the previous round again as a reference and clearly state what needs to be retained and modified in this round. This clarifies the editing baseline without relying on implicit history; for character or product series, stable identity-feature descriptions should also be retained to reduce unnecessary changes.
The endpoint returns an image URL along with task_id and trace_id, making it easy to associate tasks with results. For batch asset workflows, use async and callback_url to handle completion notifications, then send images through screening, review, and archiving steps; before formal release, the subject details and text still need to be checked.
Choose a development approach or application suited to this model and follow its setup guide.
↗Python SDK Integration Guide
SDK installation, authentication and examples↗TypeScript SDK Integration Guide
SDK installation, authentication and examples
↗
↗
↗Pricing
$0.0172· per generation
You pay$13,584for 142,800 credits · $0.0951 / credits46% OFF
You're billed in credits. Pick a top-up tier to see its per-credit rate — the larger the top-up, the cheaper each credit.
This is only a basic call example. See the full docs for more parameters and advanced usage.
View Docs