Midjourney Alpha Introduces AI-Powered Web-Based Image Creation

Midjourney Alpha Introduces AI-Powered Web-Based Image Creation

Join our daily and weekly newsletters to get the latest updates and exclusive content on top AI coverage.

Midjourney, a leading AI image generator with over 17.5 million users on its Discord server, has started testing an “alpha” version of its new website, midjourney.com. This site allows users to create images directly, bypassing the need for Discord.

The Midjourney Alpha website, which includes a feature for generating images, was noticed yesterday by dedicated users like Nick St. Pierre (@NickFloats), who shared screenshots and videos on X.

Midjourney confirmed in its Discord announcements channel that they have begun testing the alpha version of the website for image creation. However, this feature is currently limited to users who have created over 10,000 images in Midjourney. They assured that more people would gain access in the coming month.

To check your total image generation count and see if you meet the 10,000-image threshold, log into Midjourney on Discord and use the slash command “/info” in any channel or directly message the Midjourney Discord bot to get an updated count.

Despite having only generated 8,383 images in lifetime usage and 5,041 images in Midjourney’s “relaxed” mode, I could still access and test the Midjourney Alpha website.

The website’s interface is clean and simple, featuring plenty of white space. It opens with an “Explore” tab displaying public images from other users. The top has a blank rectangular text bar labeled “Imagine” where eligible users can type in prompts. An icon on the right side of this bar, representing dials, opens a visual interface to adjust prompt settings.

This setup is a major change from the current Discord version, where users must input specific text commands. For example, instead of typing “–ar 4:3” for a landscape image, users can simply select the “Landscape” option in the prompt settings menu, which automatically applies without adding any text to the prompt.

Users can click the “Photo” icon on the left side to see a timeline of their generated images, organized by date, along with current image generation jobs. This interface notably resembles Visual Electric, an AI image generation service for professional designers launched recently. However, Midjourney is often praised for its superior image quality.

The Alpha website features a tab labeled “Likes” for all images the user has “hearted” on Discord and a search bar below the image generation field to find public images by prompt keywords. Additionally, a “sun/moon” icon toggles light and dark modes, and a “thumbs up” button lets users participate in a crowdsourced image ranking project to improve the AI model.

One limitation is that images generated on the web interface do not sync back to the Midjourney Discord Bot chat, so users must switch between interfaces.

Overall, the alpha release is impressive and will likely broaden Midjourney’s accessibility. However, users still need to sign in with their Discord credentials. Midjourney, like other AI image generation services, is currently facing lawsuits from artists claiming copyright infringement due to the use of publicly available images for training its models.

More features are expected soon, including an iOS app and organizational folders on the web. Midjourney power user St. Pierre shared that Version 6 is anticipated next week, along with many more features as the alpha phase progresses.

Stay updated with the latest news by subscribing to our daily newsletter.