Image Generation

Image generation on Cortex Canvas creates visuals from the research you've already assembled. Instead of describing an image into a blank prompt, you connect research nodes, transcripts, and competitor analysis to an image node. The visual carries the argument of your content because it's built from the same sources.

How it works

1. Create an Image node on your canvas.
2. Connect it to the research nodes you want the image to draw from.
3. Choose a quality tier and style.
4. Generate the image.

The output is grounded in your research, not pulled from a stock library or generated from a separate description.

Attaching images to articles

You can attach up to two canvas-generated images to any article in the content editor. Images are attached with image and alt tags included, ready for publishing to your CMS.

For unlimited in-article visuals, you can also generate images directly in the editor using Napkin AI. See Writing Content for details on editor-based visual generation.

Use cases

• Concept illustrations that advance the argument of your article.
• Competitor comparison visuals built from tracked positioning data.
• Data visualisations grounded in research findings.
• Diagrams and flows that explain your product or process.