⚡️ Saturday AI Sparks 🤖 - 🎨🤖 Generate Images with Stable Diffusion

Text-to-Image generation has become one of the most exciting applications of AI. With just a text prompt, you can create detailed, high-quality visuals for design, art, or creative projects.

In this post, we’ll use the Stable Diffusion API from Stability AI to generate images directly from Python.

Why Stable Diffusion?

• Creative control → the image is guided entirely by your prompt.
• Flexible → choose resolution, steps, guidance scale, and samplers.
• Accessible → available via REST API, no need for heavy local setup.
• Use cases → digital art, marketing visuals, mockups, and prototyping.

Installing Requirements

We’ll use requests to interact with the API and Pillow/matplotlib for preview.

pip install requests pillow matplotlib

Setting Up the API Key

You’ll need a free account on Stability AI to generate an API key. Keep it private.

import os

STABILITY_API_KEY = os.getenv("STABILITY_API_KEY")

if not STABILITY_API_KEY or STABILITY_API_KEY == "YOUR_API_KEY_HERE":
    raise ValueError("Please set STABILITY_API_KEY to your actual key before running.")

Defining the Prompt

The quality of results depends heavily on your prompt. Here’s an example:

prompt = "A futuristic hi-tech coding desk with holographic screens, neon lighting, ultra-detailed, cinematic digital art"
negative_prompt = "blurry, low-res, distorted, text overlay, watermark"

Sending the API Request

Stable Diffusion requires JSON input with text_prompts and generation parameters.

import requests

endpoint = "<https://api.stability.ai/v1/generation/stable-diffusion-xl-1024-v1-0/text-to-image>"

headers = {
    "Authorization": f"Bearer {STABILITY_API_KEY}",
    "Content-Type": "application/json",
    "Accept": "image/png",
}

payload = {
    "text_prompts": [
        {"text": prompt, "weight": 1},
        {"text": negative_prompt, "weight": -1}
    ],
    "cfg_scale": 7.5,
    "steps": 30,
    "width": 1024,
    "height": 1024,
    "samples": 1,
}

resp = requests.post(endpoint, headers=headers, json=payload)

if resp.status_code == 200:
    with open("sdxl_output.png", "wb") as f:
        f.write(resp.content)
    print("✅ Image saved as sdxl_output.png")
else:
    print("❌ Generation failed:", resp.text[:200])

Previewing the Result

You can display the output in a notebook or script:

from PIL import Image
import matplotlib.pyplot as plt

img = Image.open("sdxl_output.png")

plt.figure(figsize=(2, 2))  # passport-size preview
plt.imshow(img)
plt.axis("off")
plt.show()

Sample Output

With the example prompt above, the API generates an artistic hi-tech desk scene complete with neon lights and holographic screens. Results vary with each run.

Key Takeaways

• The Stable Diffusion API makes image generation as simple as a POST request.
• Prompt engineering is the key — detailed prompts produce better results.
• You can customize style, resolution, and guidance for different aesthetics.
• Useful for designers, creators, and developers experimenting with generative AI.