Creating Images with AI

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:

1. Subject: What is the main focus of the image? (e.g., “A modern office building”, “A team of diverse professionals”)
2. Style: What is the artistic style? (e.g., “Photorealistic”, “Flat vector illustration”, “Oil painting”, “3D render”)
3. Mood/Lighting: What is the atmosphere? (e.g., “Bright and sunny”, “Moody and cinematic”, “Soft pastel colors”)
4. Composition: How is the image framed? (e.g., “Wide angle”, “Close-up”, “Minimalist background”)

​

Example Prompts

​

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 supports multiple image generation models from OpenAI and Google. Your team administrator enables the models your organization needs and sets the default in Team Settings > Models. Image generation must also be turned on for your group in Group Management. Image models are separate from chat models. The model selector in the chat header controls which AI model powers the conversation. It does not switch image generation models. All images in your workspace use the default image model configured by your administrator.

​

OpenAI Models

• GPT-Image-2: OpenAI’s latest image model for fast, high-quality generation and editing.
• GPT-Image-1.5: Strong prompt adherence and image quality for polished creative work.
• GPT-Image-1: High-fidelity general-purpose image generation.
• GPT-Image-1 Mini: Cost-effective option balancing speed and quality for everyday visuals.

​

Google Models (Nano Banana)

• Nano Banana 2: Fast and efficient model for quick iterations and responsive creative workflows.
• Nano Banana Pro: Advanced model for richer compositions and more demanding creative tasks.
• Nano Banana: Balanced speed and quality for general-purpose image generation.

​

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 for later use. You can browse them in two places:

1. Assets tab in chat: In the chat sidebar, switch from Chats to Assets to see generated images from your conversations.
2. Generated Images page: Click Generated Images in the main navigation bar to open the full gallery with all your images across the workspace.

For details on viewing, downloading, and navigating back to source conversations, see the Generated Images section in the platform documentation.