JUHE API Marketplace

Nano Banana 2 Facade Material Studies: Generate 20 Cladding Variants from One Prompt at 2K Resolution

6 min read
By Chloe Anderson

Start experimenting with batch generation of facade material variants using WisGate’s cost-effective API to accelerate your architectural visualization workflow. WisGate provides unified access to advanced AI models that let developers and design professionals generate high-quality images efficiently and affordably.

Overview of Nano Banana 2 Facade Material Study Goals

The Nano Banana 2 facade project focuses on creating multiple cladding material variants quickly to support client decision-making in architectural visualization. Batch generation workflows allow generating 20 unique cladding options from a single text prompt, ensuring consistent style while exploring material diversity. This method reduces manual image editing and accelerates presentation-level asset creation.

Generating multiple facade cladding variants is essential for evaluating finishes, textures, and lighting conditions before physical mockups or costly renderings. The batch process enables architects and designers to deliver diverse, high-resolution options with minimal overhead. Using WisGate’s AI image API, it becomes possible to programmatically generate these 2K resolution images in a controlled, repeatable manner.

This project targets presentation-quality outputs where each image is suitable for client review and selection. The ability to run 20 image generations in one session with a stable ~20-second turnaround per image creates an efficient feedback loop.

Setting Up Batch Variant Generation with WisGate API

To start generating material studies in batch, you first need access to the WisGate API with your API credentials. The setup involves the following steps:

  1. Obtain your WisGate API key after signing up at https://wisgate.ai/.
  2. Prepare your development environment with tools like curl or a HTTP client.
  3. Define your workspace or script to send multiple prompt variations to the API.

Batch workflows depend on programmatically tweaking prompt parameters such as material type, surface finish, or lighting to produce distinct images. WisGate’s unified API simplifies this with a consistent endpoint.

API Endpoint and Model Selection Details

The core endpoint for this project is:

POST https://wisgate.ai/v1beta/models/gemini-3-pro-image-preview:generateContent

This endpoint uses the gemini-3-pro-image-preview model capable of image generation at 2K resolution with a fixed 1:1 aspect ratio. It is well-suited for producing detailed facade cladding variants.

Request headers require your API key:

-H "x-goog-api-key: $WISDOM_GATE_KEY"
-H "Content-Type: application/json"

The API returns base64-encoded images inside the JSON response which you can then decode and save.

Crafting the Core Prompt for Material Variants

The success of batch variant generation hinges on constructing a clear, descriptive base prompt. This prompt specifies the architectural facade features and guides the variations.

For example, your base prompt could be:

"Facade cladding material study featuring brushed aluminum, terracotta, and natural stone finishes under soft daylight. Each variant emphasizes texture detail, reflectivity, and realistic shadowing."

Vary parameters by adjusting:

  • Material type (e.g., from aluminum to wood)
  • Surface finish (matte, glossy, rough)
  • Lighting conditions (direct sunlight, overcast)

Including such details ensures each generated image has distinct visual cues while maintaining an overall consistent style ideal for comparative presentation.

Technical Specs and Parameters for 2K Resolution Output

Setting technical generation parameters correctly is critical for achieving presentation-quality results. Key image specs include:

  • Aspect Ratio: Fixed at 1:1 for square images, ideal for modular cladding presentations.
  • Image Size: 2K resolution (~2048x2048 pixels) balances clarity and generation time.
  • Generation Time: Approximately 20 seconds per image for base64 images between 0.5K and 4K resolution.

Configuring these parameters on the WisGate API call looks like this:

"generationConfig": {
  "responseModalities": ["TEXT", "IMAGE"],
  "imageConfig": {
    "aspectRatio": "1:1",
    "imageSize": "2K"
  }
}

Consistent output timing promotes efficient batch creation workflows, letting you scale from 1 to 20 images without unpredictable delays. The 2K resolution ensures facade details like grain, pattern, or reflectiveness appear clearly for client review.

Executing and Automating Generation of 20 Cladding Variants

Automation scripts can loop through prompt variants to call the WisGate API and store the resulting images. Here’s a cURL snippet illustrating a single request:

curl -s -X POST \
  "https://wisgate.ai/v1beta/models/gemini-3-pro-image-preview:generateContent" \
  -H "x-goog-api-key: $WISDOM_GATE_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [{
      "parts": [{
        "text": "Facade cladding study with brushed aluminum finish under soft daylight lighting."
      }]
    }],
    "generationConfig": {
      "responseModalities": ["TEXT", "IMAGE"],
      "imageConfig": {
        "aspectRatio": "1:1",
        "imageSize": "2K"
      }
    }
  }' | jq -r '.candidates[0].content.parts[] | select(.inlineData) | .inlineData.data' | base64 --decode > cladding_variant.png

To generate 20 variants, you can script multiple JSON payloads varying the prompt text for materials, finishes, or lighting conditions. Collect and decode each base64 response image for your presentation assets.

Cost and Efficiency Comparison

Generating 20 high-resolution images at scale requires controlling API costs and ensuring predictable performance.

WisGate offers the gemini-3-pro-image-preview model at a cost of $0.058 USD per image, compared to the official rate of $0.068 USD per image. This roughly 15% cost saving accumulates meaningfully over large batch runs.

Moreover, WisGate delivers a consistent image generation time of ~20 seconds per image across resolutions from 0.5K up to 4K base64 outputs. This stability supports reliable scheduling and resource allocation.

ProviderCost per Image (USD)Generation Time
WisGate API0.058~20 seconds
Official API0.068Variable

Adopting WisGate’s API for facade material studies can reduce time-to-delivery and overall project costs while providing professional-quality outputs.

Presenting and Utilizing Generated Material Variants

Once generated, the 20 facade cladding variants can be assembled into architectural presentations. The uniform 2K resolution square images facilitate grid layouts and easy side-by-side evaluation.

You can integrate these images into PDFs, web galleries, or interactive slides to showcase material options to clients or stakeholders. The variety of finishes and lighting conditions displayed helps clarify design decisions and supports faster approval cycles.

Variations derived from prompt parameters—material type, surface finish, and lighting—highlight subtle distinctions useful for understanding aesthetic and functional impacts.

The presentation-quality output from WisGate API ensures that no post-processing is required unless specific branding or annotations are needed.

For interactive experimentation, the WisGate AI Studio at https://wisgate.ai/studio/image allows designers to preview and iterate on prompt ideas before batch generation.

Conclusion and Next Steps with WisGate Studio and API

Generating multiple facade material variants at 2K resolution using WisGate’s gemini-3-pro-image-preview API offers a streamlined, cost-effective workflow for architectural visualization projects like Nano Banana 2.

This approach enables the creation of 20 client-ready cladding options from a single prompt session, balancing image quality, generation speed (~20 seconds per image), and competitive pricing ($0.058/image).

To explore these batch workflows further, try WisGate AI Studio for on-demand image testing or integrate the API using the example endpoint and model information:

Whether you are a developer or design professional, WisGate’s image generation platform supports accelerated production of presentation-quality architectural material studies, helping you build faster and spend less with one unified API.

Nano Banana 2 Facade Material Studies: Generate 20 Cladding Variants from One Prompt at 2K Resolution | JuheAPI