Generate images with Dream Lab

Dream Lab is Canva’s AI image generator for creating detailed visuals from text prompts. You can generate multiple image options at once. You can use a reference image to guide results. You can refine outputs with follow-up edits. Some Dream Lab features use your AI usage allowance when applied.

• Enter a prompt (a short description of what you want to create).
• Dream Lab generates four image options for each prompt.
• Select a result to use in a design. Or refine it with a follow-up instruction, such as “Make the car blue.”
• (Optional) Add a reference image to guide the style or composition. Using reference images may use AI usage, depending on how they’re applied.
• (Optional) Select Edit image to make specific changes to a generated result.
• (Optional) Use Boost resolution to make your generated image clearer and more detailed. This feature uses AI usage.

Dream Lab offers different quality tiers for creating images from text. Higher tiers use more AI usage but produce better quality results. Generation times may be slower with higher tiers.

You can select a quality tier when:

• No style is selected
• No image is attached

To select a quality tier:

1. Enter your text prompt.
2. Select the tier option before generating.
3. Choose your preferred tier based on quality and credit cost.
4. Select Generate.

You can use Dream Lab to edit images you've generated. This gives you more control over your results.

1. Generate an image in Dream Lab.
2. Select the image you want to edit.
3. Select Edit image.
4. Describe the changes you want to make.
5. Select Generate.

Boost resolution makes your Dream Lab images clearer and more detailed. Unlike Upscale, which improves pixelated and low-resolution images, Boost resolution increases the resolution of AI-generated images.

Here's how to do it:

1. Generate an image in Dream Lab.
2. Select the image you want to enhance.
3. Select Boost resolution.

Dream Lab keeps a history of the images you’ve generated. To find previous results:

• Open Dream Lab
• Select History to see your past images.

• Be specific. Include details like style, colours, lighting, and layout.
• Describe the subject and composition directly (e.g., “top-down flat lay,” “wide-angle landscape,” “portrait with soft studio lighting”).
• Generate variations. Then refine your prompt based on what you like or dislike in the results.
• If using a reference image, try different reference options (if available) to better control how closely the result follows it.

• Unstable internet connection
• Prompt may not meet Canva’s content guidelines

1. Check your internet connection.
2. Try generating again.
3. Review your prompt and remove any potentially restricted or inappropriate content.

• You’re not accessing Dream Lab from the Homepage
• You’re trying to find it inside the editor

You may be looking at the wrong place for your Dream Lab history.

Open Dream Lab and check your history there.

• Supports all languages available on Canva, except for users in canva.cn.
• Generates up to four images per prompt.
• Results depend on the quality and detail of your prompt.
• Accessible from the Homepage (not inside the editor).
• You can choose a quality tier only when you generate from text without a style and without attaching an image.

Yes. Admins, teachers, and brand designers can manage access for their team or class.

There’s no easy answer to this since the treatment of AI-generated works under copyright law varies country by country.

You own the images and videos you create with Dream Lab, as long as you follow our Terms. However, whether AI-generated images are copyrightable depends on the jurisdiction. Under our Terms, you allow us to host the images you generate. This helps us provide the Dream Lab service to you.

Similarly, if you use these apps to create recognisable characters or brands, these may be protected by copyright or trademarks. If you plan to use these images, please get permission from the rights owner. We can't give you this permission.

Our values define how we build our products, treat our community, and impact the world. We want to ensure our community feels safe when creating and interacting with content on our platform.

We use automated checks on prompts and images to block or flag content that may be inappropriate. These measures help keep Dream Lab safe for everyone. You can also report unsafe images to us.

We do not tolerate any use of these apps that violates our Terms or Acceptable Use Policy. Read through this page to learn how to use Dream Lab and other Magic Studio features safely and legally.

You can report an image using the Report button in the app. Or, get in touch with us to provide feedback at any time.

Some features, such as reference images, Boost resolution, and Generating images consumes AI usage allowance.. You’ll be notified whenever a feature requires AI usage.

AI-generated content must follow Canva's Acceptable Use Policy. We block prompts that request harmful, illegal, or inappropriate content.

Generated images include a visible label showing that AI created them. This helps make AI-generated content clear and honest.