Skip to navigation

Image Generation

View as Markdown

Create stunning, high-quality images from text descriptions using NeuraAI’s advanced image generation models.

Basic Usage

Generate images with a simple text prompt:

from openai import OpenAI
client = OpenAI(
base_url="https://api.neura-ai.app/v1"
)
response = client.images.generate(
model="dall-e-3",
prompt="A futuristic city at sunset with flying cars and neon lights"
)
print(response.data[0].url)

Prompt Enhancement

Enable automatic prompt enhancement to get better results:

response = client.images.generate(
model="dall-e-3",
prompt="A cozy coffee shop interior",
extra_body={"enhance": True}
)
print(response.data[0].url)

When enhance is enabled, the API automatically improves your prompt by adding artistic details, style specifications, and quality modifiers.

Image Size

Some models support custom image dimensions:

response = client.images.generate(
model="dall-e-3",
prompt="A majestic mountain landscape",
size="1024x1024" # Check model documentation for supported sizes
)

Note: Not all models support the size parameter. Refer to the Models page to verify which models offer size customization.

Supported Sizes

Common size options (model-dependent):

  • 1024x1024 - Square format
  • 1792x1024 - Landscape
  • 1024x1792 - Portrait

Advanced Examples

Detailed Art Generation

response = client.images.generate(
model="dall-e-3",
prompt="""
An oil painting of a serene Japanese garden during cherry blossom season.
Include a wooden bridge over a koi pond, stone lanterns, and Mount Fuji
in the background. Style: traditional impressionism with soft brush strokes.
""",
extra_body={"enhance": True}
)

Product Visualization

response = client.images.generate(
model="dall-e-3",
prompt="""
A sleek, modern smartwatch on a marble surface with soft lighting.
Professional product photography style, studio quality, high detail.
""",
size="1024x1024"
)

Character Design

response = client.images.generate(
model="dall-e-3",
prompt="""
A cyberpunk character portrait: female hacker with neon blue hair,
augmented reality glasses, leather jacket with glowing circuit patterns.
Dark background with holographic displays. Digital art style.
""",
extra_body={"enhance": True}
)

Architectural Concept

response = client.images.generate(
model="dall-e-3",
prompt="""
Futuristic eco-friendly skyscraper with vertical gardens, solar panels,
and organic curved architecture. Surrounded by lush greenery.
Architectural rendering style.
""",
size="1024x1792"
)

Working with Generated Images

Downloading Images

import requests
from PIL import Image
from io import BytesIO
response = client.images.generate(
model="dall-e-3",
prompt="A peaceful forest scene with sunlight filtering through trees"
)
# Get the image URL
image_url = response.data[0].url
# Download and save
img_response = requests.get(image_url)
img = Image.open(BytesIO(img_response.content))
img.save("generated_image.png")

Multiple Variations

Generate multiple images from the same prompt:

for i in range(3):
response = client.images.generate(
model="dall-e-3",
prompt="A mystical wizard casting a spell"
)
print(f"Variation {i+1}: {response.data[0].url}")

Prompt Writing Tips

Be Specific

Instead of: “A dog” Use: “A golden retriever puppy playing in autumn leaves, warm afternoon sunlight”

Include Style

Specify artistic style for better results:

  • “Watercolor painting of…”
  • “Photorealistic render of…”
  • “Minimalist line drawing of…”
  • “3D rendered…”

Describe Composition

Guide the layout and framing:

  • “Close-up portrait of…”
  • “Wide-angle landscape showing…”
  • “Bird’s eye view of…”
  • “Centered composition with…”

Set the Mood

Add atmosphere and lighting:

  • “Dramatic lighting with strong shadows”
  • “Soft, dreamy atmosphere”
  • “Vibrant and energetic”
  • “Moody and mysterious”

Best Practices

  • Use descriptive, detailed prompts for better results
  • Enable prompt enhancement for automatic improvements
  • Experiment with different artistic styles
  • Specify desired mood, lighting, and composition
  • Check model capabilities before using size parameter
  • Test multiple generations for the best output
  • Combine subjects, styles, and details thoughtfully

Common Use Cases

  • Marketing materials and advertisements
  • Social media content
  • Concept art and design mockups
  • Book covers and illustrations
  • Product visualizations
  • Educational materials
  • Personal creative projects