NanoBanana MCP
Generate and edit images using Nano Banana in the AI Agent through MCP.
Based on Google Gemini 2.5 Flash/3 Pro Image models, enabling high-quality image generation and editing through a simple REST API. Supports multiple scenarios including text-to-image, multi-image editing, virtual try-on, and product background replacement.
Nano Banana is a unified API wrapper based on Google Gemini image generation models. It simplifies and optimizes the Gemini 2.5 Flash Image and Gemini 3 Pro Image models, offering a more user-friendly interface and lower pricing.
Ace Data Cloud offers lower prices than the official service (approximately 1/3 of the official price), along with higher concurrency support, dedicated customer service, and 24/7 stability assurance. No need to handle complex Google Cloud configuration—ready to use right away.
Text-to-image and multi-image editing, covering the complete image creation workflow
Enter a detailed text description (subject, atmosphere, lighting, camera) to generate professional-quality, high-texture images. Ideal for brand visuals, product posters, cover images, and more.
Upload multiple images (such as a portrait + clothing), and AI will intelligently blend them. Supports advanced editing operations such as virtual try-on, product placement, and scene compositing.
Provide a portrait photo and a clothing PNG to automatically create the outfit effect. Perfect for e-commerce apparel displays and quickly expanding livestream images.
Place product images on white backgrounds into real scenes (wooden tables, grass, bathrooms, desks) to make product detail pages richer and more full of life.
Change materials, colors, and styles. Swap a bag for leather, a beverage for a different flavor, or shoes for a new colorway—making visual SKU experimentation easier.
Set callback_url, and results are automatically pushed when the task is complete. No polling required; ideal for batch production scenarios.
The following images were actually generated by the Nano Banana API—what you see is what you get
nano-banana · 3:2 · Text-to-image
nano-banana · 16:9 · Text-to-image
nano-banana-pro · 1:1 · Text-to-image
nano-banana · 3:4 · Text-to-image
💡 All images were generated through a single API call, with no post-processing
curl -X POST https://api.acedata.cloud/nano-banana/images \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"action": "generate",
"prompt": "白瓷咖啡杯,晨光窗台,85mm镜头,浅景深,柔和高光,治愈干净",
"model": "nano-banana",
"aspect_ratio": "3:2"
}'
{
"success": true,
"task_id": "c1075552-a527-4a70-8d98-ef783f4087c1",
"trace_id": "3352b744-d17e-4d01-b1ef-a8c8e2ca3d2e",
"data": [{
"image_url": "https://cdn.acedata.cloud/d635720766.png"
}]
}
A simple REST API using Bearer Token authentication. Generate your first AI image with a single request.
Register with Ace Data Cloud and get your Bearer Token from the console
Send a request to /nano-banana/images with a prompt and parameters
Receive a permanently accessible image URL—ready for production use immediately
From e-commerce content to brand creativity—Nano Banana covers these scenarios
Take one styled photo with an influencer, then batch-generate "outfit images" using different clothing PNGs to quickly expand your store assets
Use a single product image on a white background to create images in multiple settings such as wooden tables, grass, or bathrooms, making detail pages richer and more persuasive
Quickly generate brand key visuals, campaign posters, and cover images. Switch the same core visual to different styles or holiday themes with one click
Integrate with MCP-compatible AI assistants through MCP Server for natural-language image generation
Nano Banana is more than simple text-to-image generation—it can also handle these advanced tasks
Convert anime characters/portraits into commercialized figurine models, complete with packaging box displays
Showcase a product's internal structure, with metal parts and electronic components disassembled and floating for technical aesthetics
Restore scratched and damaged old black-and-white photos, enhance clarity, and add natural color
Naturally blend multiple assets such as models, products, and backgrounds into a single high-quality image
From registration to generating your first AI image, it takes less than 3 minutes
Create a free account on Ace Data Cloud. Generate your Bearer Token from the API management console.
Send a POST request with a text prompt to /nano-banana/images. You can use an SDK, cURL, or any HTTP client.
Embed the API into your application. Use Webhooks for asynchronous processing and scale confidently to production.
Compare with Google Official and see our advantages
| Feature | Ace Data Cloud | Google Official |
|---|---|---|
| Price | As low as $0.0133 / request | $0.039 Interactive / $0.02 Batch |
| Getting Started Requirements | ✓ Ready to use instantly | ✗ Google Cloud configuration required |
| Concurrency Support | ✓ High concurrency guaranteed | ✗ Low default quota |
| Webhook Callbacks | ✓ | Partially supported |
| Task Polling | ✓ Free | — |
| MCP Server | ✓ | ✗ |
| Dedicated Customer Support | ✓ | ✗ |
| 24/7 Monitoring | ✓ | Partially supported |
Choose the right model based on your quality and speed needs
Based on Gemini 2.5 Flash Image. Faster generation, lower cost, ideal for batch content production and prototype validation.
Based on Gemini 3.1 Flash Lite Image. Supports 1K resolution only, with fast image generation in 3–5 seconds.
Based on Gemini 3.1 Flash Image. Pro-level quality + Flash speed, an excellent value choice.
A new-generation efficient image model supporting 1K, 2K, and 4K text-to-image generation and image editing.
Based on Gemini 3 Pro Image. Highest-quality output, supporting 2K/4K resolutions. Ideal for flagship products and demanding use cases.
Transparent pay-as-you-go pricing. No subscription fees. No hidden costs. Pay only for what you use.
nano-banana model (Gemini 2.5 Flash Image)
Save up to 27% with volume packages
Volume discounts for high-usage teams
Everything you need to know about using the Nano Banana API
Nano Banana is an alias wrapper for Google Gemini image generation models. nano-banana corresponds to Gemini 2.5 Flash Image, nano-banana-2-lite corresponds to Gemini 3.1 Flash Lite Image (1K only, fast image generation), nano-banana-2 corresponds to Gemini 3.1 Flash Image, nano-banana-2.1 corresponds to Gemini Nano Banana 2.1 (supports 1K, 2K, and 4K generation and editing), and nano-banana-pro corresponds to Gemini 3 Pro Image.
Pay as you go, with no subscription fee. nano-banana and nano-banana-2-lite models start at $0.0133 per generation, nano-banana-2 models start at $0.0267 per generation, nano-banana-2.1 models start at $0.0190 per generation, and nano-banana-pro models start at $0.0333 per generation. Task status polling is always free; please refer to the real-time pricing page for specific prices.
Two operation modes: Generate (text-to-image) and Edit (multi-image editing). Generate supports 7 aspect ratios (1:1, 3:2, 2:3, 16:9, 9:16, 4:3, 3:4), while the Pro model also supports 1K/2K/4K resolution options. Edit mode supports uploading up to 10 images for combined editing.
Two options: (1) Use callback_url—set a Webhook URL, and we will POST the result when complete. (2) Poll via /nano-banana/tasks—this endpoint is free and lets you check the status at any time.
Yes. The returned image_url is stored on our CDN and remains permanently accessible, ready for direct use in production environments.
Yes! We provide an MCP Server (mcp-nano-banana) that can integrate with MCP-compatible AI assistants. Your AI Agent can generate and edit images through natural language.
Recommended structure: subject + atmosphere + lighting + lens + texture keywords. For example: "White porcelain coffee cup, morning-lit windowsill, 85mm portrait lens, shallow depth of field, soft highlights, soothing, clean, leisurely." The more specific your description, the better the generated result.
Edit mode requires passing 1-10 images through the image_urls parameter (HTTP/HTTPS URLs or Base64). Each image can be up to 10MB. The Prompt should describe how to process these images, for example, "Make this person wear this T-shirt."
Explore our complete AI API suite, covering images, videos, music, and more
As low as $0.0133 per call—lower prices than the official service, a simpler interface, and higher concurrency support. Pay as you go—no subscription fees, no commitments.