You Can Now Edit AI Images On Midjourney Website

Merely weeks after Midjourney released the V6.1 image model, they announced three new updates, though this time there’s no shiny new image model. Instead, they’ve rolled out improvements on the Alpha website that greatly improves the way you use the image generator.

The updates include the following:

• Expanded user access: Users who have made at least 10 images on Midjourney now have access to the Midjourney Alpha website.
• Room syncing: Midjourney is now syncing images and messages between rooms on the web and channels on the Discord server. This means that messages and images in certain rooms/channels are reflected back onto Discord and vice-versa.
• Image editor on the website: Users can now edit Midjourney images directly on the website, cutting out the need to rely solely on Discord.

In this article, let’s explore the features of the image editor tool, and how it compares to the Discord version. Let’s get started.

The New Web Editor Tool

The new image editing tool combines various editing actions like reframe, repaint, vary region, pan, and zoom into a single interface.

This is what the user interface looks like:

In case you didn’t know it yet, editing Midjourney images has already been around for years on the Discord server. For anyone who’s spent time trying to edit images on Discord, this is a breath of fresh air. The old method on Discord was functional but far from ideal — it felt more like a workaround than a feature.

Having a dedicated image editor built directly into the website is a huge relief, especially for users who frequently need to tweak their images.

Main Features

In the “Create Actions” section of the web editor, you’ll notice two new buttons: Rerun and Editor.

• Rerun: This button allows you to run the same prompt again.
• Editor: Edit specific parts of the image, with or without a prompt change.
• Remix: Edit your prompt text and parameters after generating an image. It can help you change the setting or lighting of an image, evolve a subject, or achieve tricky compositions.

The other editing controls are pretty self explanatory.

Discord VS Web Image Editor

To understand how the new web image editor stacks up, let’s take a closer look at how things used to work on Discord. After you generate and upscale an image, you have the option to click “Vary (Region)” to edit a portion of the image.

A pop-up window opens, allowing you to select the region to modify using either a square or a lasso selection tool.

This method worked, but it was clunky. The user experience wasn’t exactly smooth, and it often felt like you were fighting against the interface to make simple changes.

The web editor, by contrast, offers brush control, which is far more intuitive. You can adjust the brush size and easily switch between erasing or restoring parts of the selection. This alone makes the web editor feel more like a professional tool than the Discord version.

Next, add the description of the modified image you want to see. In the example above, I wanted to replace the loaf of bread with a big piece of broccoli, so here’s the new prompt I used.

Prompt: kitten hugs a fat chonky grey cat laying on its back a big broccoli

The pop-up window will close, and the new set of images will be displayed on the chat history screen.

After the image is modified, you have to reopen the editing pop-up if you want to make further changes. This process can quickly become tedious, especially if you’re making multiple edits.

The web editor simplifies this workflow. After clicking the submit button, the modified images appear under the Create tab rather than forcing you to go back and forth between different screens. The less friction there is in the editing process, the more likely users are to experiment and create.

Buut… this is also not good in terms of user experience. Why can’t they just display the image results on the image editor dashboard?

Another trick you can do on the web editor is to use the scale control to outpaint the image. For example, if I wanted to see more of the background full body of the kitten, I’d scale it down to 70% and move the image to the left side.

Click on the Submit button and this is the final result:

Awesome. Look how cute that cat is.

Okay, that’s about it. What do you think about this new image editor on the Midjourney website? Is it time to ditch the Discord version?

Final Thoughts

Overall, I am impressed with how well the image editor works and how helpful it is for those who often edit their AI images. One minor area for improvement is how the results are displayed, perhaps a pop-up window to show the results would be a huge improvement in the usage experience.

With these updates, I find myself using the Discord version less and less. The ease of use and the seamless workflow for generating and modifying images on the website make it far more practical, even though it’s still in the alpha phase.

If you want to learn more about the features of Midjourney and how to use them, check out their official documentation here. Also, make sure to follow the publication for more updates in the generative AI space.

‍