Ideogram has become the most recent artificial intelligence platform to incorporate a canvas approach to generating content. Its latest upgrade comes in the form of an editor that allows users to paint any element of an image, layer images, and generate new images.

There seems to be a growing trend among AI companies to find new ways for users to interact with their models beyond simple text or chat input; Anthropic led some of this charge with Claude's Artifacts, and then OpenAI has been working on its own text and code editing canvas picked it up.

I spent some time experimenting with the Ideogram Canvas and was impressed with its flexibility: unlike other web image editor approaches that use AI, Ideogram has a layering feature, where each new generation or change is layered on top of the original generation. This makes it very easy to revert and compare changes between one version and the next.

This editor is based on concepts that Midjourney presented in their own editor, such as expanding the canvas, painting around certain elements, and making changes with simple text prompts. I have used both and find Ideogram to be the better implementation.

The canvas can be used as a stand-alone image generation interface or as a place to submit previously generated images. Compared to a text input field, one can more easily review the image and make changes.

A spokesperson explains: “Ideogram Canvas is an endless creative board for organizing, generating, and editing images. Images can be uploaded or generated and then seamlessly edited or enhanced with Ideogram's industry-leading in-painting and out-painting tools."

The ability to upload images and edit/enhance them using Ideogram is a powerful feature not often available on other platforms. Apparently Midjourney is working on making this functionality readily available in the editor, but so far it is not possible.

This is especially powerful in the realm of graphic design, as Ideogram's text rendering capabilities are different from any other model I have tried. For example, when I uploaded an image I had created, I was able to add text to the frame in seconds.

You can also give it your own images and remix them. For example, I could change the color of my cat or place my cat in space. I don't actually have a cat, but I have found that most AI tools are better at generating images of cats than dogs.

One of my favorite features of Ideogram Canvas is the ability to rename the canvas you are working on or create a new canvas. This creates a “project” style system of AI images.

In fact, the core platform of Ideogram has not changed. It has been redesigned to look more like the Midjourney app, with generations displayed on the side and canvas buttons on the side, but the functionality is the same.

Other new features I noticed in addition to the design changes and canvas are the new animation customization in addition to the existing design, 3D, and realistic options. This is a tool that further refines the prompts and brings them closer to the animated image.

Note: If you use Canvas and, like me, prefer a tidy archive of previously created generations, any changes you make to an image in Canvas will appear as a new generation in your listing.