Image Generation

When the AI agent generates an eLearning course, it also creates images for each slide — illustrations, icons, decorative visuals, and more. You don't need to prepare any images upfront. The agent reads your content and the template's design instructions, decides what visuals each slide needs, and generates them automatically as part of the course creation process.

Generated Images as Placeholders

In many production workflows, you'll eventually want to replace some or all of the generated images with your own — brand photography, approved illustrations, or assets from your design team. AI-generated images work well as placeholders during the first iteration. They give you a complete, visually coherent draft to review and share, without waiting for final assets. When you're ready, simply upload your own images and ask the agent to swap them in.

This approach lets you focus on content structure and pedagogical flow early on, while deferring visual polish to a later stage.

How It Works

You don't need to write detailed image prompts yourself. Describe what you want in your own words — the AI agent takes your description, combines it with the context of the slide and the template's design guidelines, and constructs a high-quality prompt under the hood when it calls the image generation model. The result is an image that fits the content and visual style of your course.

During Course Generation

When you ask the agent to generate a full course, images are created alongside the content automatically. The agent decides what each slide needs based on the topic, the template structure, and any design instructions.

Requesting Changes

You can always ask for changes to specific images after the initial generation:

Replace the image on the "Key Concepts" slide with something 
that shows a person interacting with a touchscreen device.

The illustration on the safety page should feel more serious 
and professional — less cartoon-like.

Uploading Your Own Images

To replace a generated image with your own, upload the file in the chat and tell the agent where to use it:

Use the attached photo as the header image on the introduction slide.

Aspect Ratios

Choose the right dimensions for your use case:

You can specify the ratio in your prompt:

Create a 16:9 hero image for the introduction slide showing 
a collaborative workplace setting.

Style Guidelines

Be Descriptive

The more context you give, the better the result. You don't need to use technical prompt language — just describe what you're looking for in plain words.

Good prompt:

I'd like a warm, friendly illustration for the onboarding slide. 
It should show a new employee being welcomed by colleagues in a 
modern office. Keep the style clean and professional, matching 
our brand colors (blue and white).

Less effective:

Onboarding image

Specify the Style

Common styles you can request:

• Flat design — Modern, minimalist, solid colors
• Isometric — 3D-like with consistent angles
• Watercolor — Soft, artistic, approachable
• Line art — Clean outlines, professional
• Photorealistic — Lifelike images
• Cartoon/Illustrated — Friendly, engaging

Include Context

Tell the agent how the image will be used so it can make better decisions:

This image will be the opening visual for a compliance training 
course aimed at factory floor workers. It should feel approachable 
but convey the seriousness of workplace safety.

Supported Model

Impact Learning uses Gemini 3.1 Flash Image for all image generation. It offers fast generation with good quality, making it well suited for iterating on illustrations and visuals during course creation.

Image Credits

Image generation uses credits from your plan. Each generated image costs approximately 1 credit. See Plans and Credits for details.