Omnifact’s Image Generation feature isn’t just for fun—it’s a powerful tool for accelerating creative workflows. Whether you need mockups for a presentation, unique visuals for social media, or concept art for brainstorming, the AI assistant can help you visualize your ideas in seconds.
This guide will help you move beyond basic requests and start creating high-quality, professional images tailored to your specific needs.
When to Use Image Generation
AI image generation shines when you need speed and flexibility. Instead of searching for stock photos or waiting for design resources, you can generate custom visuals on demand.
Common professional use cases include:
- Presentations: Create custom illustrations or diagrams to visualize complex concepts.
- Marketing & Social Media: Generate unique, on-brand visuals for posts, ads, or blog headers.
- UI/UX Mockups: Quickly visualize layout ideas or icon styles before starting high-fidelity design.
- Brainstorming: Explore different visual directions or moods for a project.
- Storyboarding: Visualize scenes or sequences for video or narrative projects.
Writing Effective Prompts
The key to getting the image you want is a clear, descriptive prompt. Think of yourself as an art director giving instructions to an illustrator.
A strong prompt typically includes these four elements:
- Subject: What is the main focus of the image? (e.g., “A modern office building”, “A team of diverse professionals”)
- Style: What is the artistic style? (e.g., “Photorealistic”, “Flat vector illustration”, “Oil painting”, “3D render”)
- Mood/Lighting: What is the atmosphere? (e.g., “Bright and sunny”, “Moody and cinematic”, “Soft pastel colors”)
- Composition: How is the image framed? (e.g., “Wide angle”, “Close-up”, “Minimalist background”)
Pro Tip: Be specific but leave room for creativity. Instead of just “a dog,” try “a golden retriever sitting on a porch at sunset, warm lighting, photorealistic style.”
Example Prompts
| Goal | Prompt Structure | Example |
|---|
| Icon | [Object] icon, [Style], [Color], [Background] | ”A rocket ship icon, flat vector style, blue and white, minimalist, white background.” |
| Photo | [Subject] doing [Action], [Location], [Lighting], [Camera Angle] | ”A diverse team collaborating around a whiteboard in a modern glass office, natural sunlight, candid shot, eye-level angle.” |
| Illustration | [Scene description], [Art Style], [Mood] | ”A futuristic city with flying cars, cyberpunk style, neon lights, rainy atmosphere, digital art.” |
Using Reference Images
Sometimes it’s easier to show than to tell. You can attach an existing image to guide the AI’s generation. This is perfect for:
- Style Transfer: Making a photo look like a sketch.
- Variations: Generating similar images with different details.
- Editing: Changing specific elements (like background color) while keeping the main subject.
For instructions on how to upload images, see Attaching Documents.
Common Editing Workflows
- “Turn this into…”: Upload a rough sketch or a simple icon and ask the AI to render it in a specific style (e.g., “3D”, “Photorealistic”).
- “Keep the composition but…”: If you like the layout of an image but want different subjects, ask for that. “Keep the composition of this image, but replace the office setting with a coffee shop.”
- “Make it look like…”: Upload a reference for the style you want (e.g., a specific artist or mood) and ask the AI to generate a new subject in that style.
Choosing the Right Model
Omnifact offers different image models optimized for various needs. Selecting the right model can significantly impact your results.
- GPT-Image-1 (DALL-E 3): Best for following complex instructions and adding specific text within images. It excels at understanding nuanced prompts.
- Nano Banana: A faster model, great for quick iterations and stylized illustrations.
- Nano Banana Pro: A higher-quality version of Nano Banana, offering better detail and realism for photographic styles.
You can switch models using the model selector dropdown in the chat header (if enabled by your workspace admin).
Iterating and Refining
It’s rare to get the perfect image on the first try. Use the chat interface to refine your results iteratively.
- Adjust the Prompt: If the style is wrong, add more style keywords (e.g., “Make it more abstract”).
- Add/Remove Details: “Add a laptop to the table” or “Remove the clouds.”
- Change Aspect Ratio: “Make it landscape” or “Generate a square version.”
Managing Your Images
Every image you generate is automatically saved to your Generated Images Gallery. This creates a personal library of assets you can revisit later.
For details on how to access, filter, and download your images, see the Generated Assets section in the platform documentation.
