Image generation and editing

Image generation is built right into your chat experience so you can create images without leaving the Juma app. Just ask the AI to create an image, and it will generate one based on your description.

How to generate an image

To create an image, simply ask the AI in a chat. Be specific in your prompt to get better results. For example, instead of saying "create a workspace image," you might say "create a hero image for a marketing agency website. Show a modern minimalist workspace with a MacBook displaying colorful data dashboards."

The AI will generate your image and display it in the chat.

Editing and refining images

You can refine your image by prompting further in the same chat. For example, you might say "make it more minimalist, remove the notebook and reduce the items on the desk." The AI will create a new version based on your feedback.

You can keep going back and forth until the image is right. All your generated images stay inside your chat for easy reference.

Finding your images in Projects

If you generate images inside a project, you can find them all in one place. Open the project, click on the Images tab, and all your generated images will be collected there.

Tips for writing good image prompts

Image prompts work a bit differently from regular prompts. Here are some tips to get better results:

Be specific about the subject. Instead of "a workspace," say "a modern minimalist workspace with a MacBook."

Describe the composition. Mention what's in the foreground, background, and the angle you want.

Specify the lighting. Warm natural lighting from a window is very different from bright overhead lighting.

Include a color palette. This helps keep your images on brand.

Mention the mood. For example, "professional but approachable" helps the AI understand the vibe you're going for.

Specify the aspect ratio. You can request square, horizontal, or vertical images depending on where you'll use them.

Using Project context for images

Image generation keeps in mind the context you give the AI inside the Project knowledge. For example, if you've added a product menu to your Project knowledge and ask the AI to generate a product image, it will use the details from that menu to create an accurate image.

This means your images can be more accurate and on-brand without having to repeat all the details in every prompt.