How to Generate Images with Gemini 3 Pro Image via the API (Updated Feb 2026)

The Gemini 3 Pro Image model (also known as "Nano Banana Pro") is Google's professional-grade image generation model, available through both the Gemini API and Vertex AI. It supports high-fidelity image generation up to 4K resolution, advanced text rendering, multi-image composition, and image editing.

What the Model Can Do

• Image generation up to 4K resolution from text prompts
• Text rendering in images with multilingual support
• Multi-image composition with up to 14 reference images
• Character consistency for up to 5 people across images
• Image editing with local masks and inpainting
• Aspect ratios: 1:1, 3:2, 2:3, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, and 21:9

Step 1 --- Get a Gemini API Key

1. Go to Google AI Studio.
2. Click Get API key and create or select a Google Cloud project.
3. Copy the key (starts with AIza...). Store it securely---never commit it to version control.

export GEMINI_API_KEY="your_key_here"

Step 2 --- Install the SDK

Python:

pip install google-generativeai

Node.js:

npm install @google/generative-ai

Step 3 --- Generate an Image (Python)

import os
import google.generativeai as genai

genai.configure(api_key=os.environ["GEMINI_API_KEY"])

model = genai.GenerativeModel("gemini-3-pro-image-preview")

prompt = "A serene Japanese garden at sunset with a koi pond and wooden bridge"
response = model.generate_content([prompt])

if response.candidates:
    image_bytes = response.candidates[0].content.parts[0].inline_data.data
    with open("image.png", "wb") as f:
        f.write(image_bytes)
    print("Saved image.png")
else:
    print("No image generated")

Step 4 --- Generate an Image (Node.js)

import fs from "fs/promises";
import { GoogleGenerativeAI } from "@google/generative-ai";

const apiKey = process.env.GEMINI_API_KEY;
const genAI = new GoogleGenerativeAI(apiKey);
const model = genAI.getGenerativeModel({ model: "gemini-3-pro-image-preview" });

async function run() {
  const prompt =
    "Product shot of matte black wireless headphones on a gray gradient background";
  const result = await model.generateContent([prompt]);
  const data =
    result.response.candidates?.[0]?.content?.parts?.[0]?.inlineData?.data;
  if (!data) {
    console.log("No image generated");
    return;
  }
  await fs.writeFile("product.png", Buffer.from(data, "base64"));
  console.log("Saved product.png");
}

run().catch(console.error);

Step 5 --- Write Better Prompts

Good image prompts are specific. Include details about:

• Subject: What is in the image
• Style: Photorealistic, illustration, watercolor, etc.
• Lighting: Natural, studio, golden hour, etc.
• Composition: Close-up, wide shot, aerial view, etc.
• Aspect ratio: Specify if you need a non-square output

Example of a detailed prompt:

A photorealistic close-up of a ceramic coffee mug on a rustic wooden table,
morning sunlight streaming through a window, shallow depth of field,
warm color palette, 16:9 aspect ratio

Step 6 --- Optimize for Cost and Quality

The model supports parameters for tuning the quality/cost trade-off:

• thinking_level: Set to low for faster, cheaper generation or high for maximum quality
• media_resolution: Choose low, medium, or high based on your needs
• Input limits: 7 MB for direct uploads, 30 MB via Cloud Storage
• Max images per prompt: 14

Important Considerations

• Follow Google's usage policies for generated content
• The model is currently in public preview---check the docs for any changes to availability or pricing
• API keys should always be stored in environment variables or secret managers, never in code
• Check ai.google.dev/pricing for current rate limits and costs

Sources

• Gemini 3 Pro Image docs: cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-pro-image
• Gemini API overview: ai.google.dev/gemini-api/docs
• Pricing and quotas: ai.google.dev/pricing
• Usage policies: ai.google.dev/policies

For more on Google's AI models, see our Gemini coverage in the directory.