Ideogram takes on DALL-E and Midjourney with its own text-to-image generator
Toronto-based AI startup Ideogram has officially launched its platform. The company specializes in generative text-to-image technology, so this is yet another text-to-image generator and competition to Midjourney, DALL-E, and Adobe Firefly.
Led by industry veterans who have previously worked on high-profile AI projects at Google Brain, UC Berkeley, and the University of Toronto, the startup aims to democratize creativity through AI. The company’s CEO, Mohammad Norouzi, is a former senior staff research scientist at Google. Speaking of Ideogram, he says that the platform’s mission is to focus on creativity with a “high standard for trust and safety.”
How does Ideogram work?
Ideogram offers capabilities similar to existing platforms like DALL-E, Midjourney, or Adobe Firefly. In other words, it allows you to convert textual prompts into images. To use the platform, you need to sign up using your Google email address. Upon the signup, you can type in your prompts and select from various rendering styles, such as anime, dark fantasy, or graffiti, to generate images.
You can choose between three different resolutions, and once you enter your prompt and select the style, you’ll get four results. From there on, you can open each and remix it, adding or removing prompts and styles.
Possibilities and limitations
Something pretty special about Ideogram is that it can generate letters. I haven’t seen either Midjourney or DALL-E do it properly, and I’ve played with both quite a bit. It does take a few attempts, but you’re almost certain to get at least one result. If you remix the one you like, you’ll easily get more properly spelled results.
But just like other text-to-image tools, Ideogram generates compelling images, but not without imperfections. There are weird hands and faces, unrealistic lines and colors, and of course – weird spelling.
Can Ideogram set new standards in creative AI?
By focusing on its user-friendly platform, Ideogram hopes to carve out its own niche in the burgeoning AI-generated art space. The founding team boasts an impressive background in AI research and application, and they believe that Ideogram can potentially surpass the capabilities of existing players like DALL-E and Midjourney. However, from what it looks like now, there’s still work to be done before this happens.
Dunja Djudjic is a multi-talented artist based in Novi Sad, Serbia. With 15 years of experience as a photographer, she specializes in capturing the beauty of nature, travel, and fine art. In addition to her photography, Dunja also expresses her creativity through writing, embroidery, and jewelry making.