How to use MidJourney Describe: Turn Any Image Into Prompt

Did you know that MidJourney has an image-to-text model? This little-known feature is called Describe, and it can be useful for when you’re struggling to find the right words for your prompts, need help describing a specific style in an image, or just want some prompt inspiration.

MidJourney recently enabled Describe directly on the website, and there are two ways to access it.

What Is the Describe Feature in MidJourney?

MidJourney’s Describe tool is essentially an image-to-text model. It looks at the visual content in an image and offers a few prompt suggestions that could help recreate a similar vibe or inspire something new.

The prompts are not meant to exactly recreate the original image.
Think of them as starting points or creative suggestions.
You can modify, use as-is, or generate new ones based on the same image.

MidJourney Describe Overview:

Platform	How to Access Describe	Prompt Output
Website	Right-click Midjourney image	4 suggestions
Website	Drag image to Imagine bar	4 suggestions
Discord	Use `/describe` command	4 suggestions
Full Editor	Use “Suggest Prompt” in editor	1 suggestion only

How to Use the Describe Feature on the MidJourney Website

Midjourney now makes it easier than ever to access Describe directly on its website. There are two main ways to use it here.

Right-Click an Image in Explore or Your Gallery

You can use Describe on images found in your Explore page or personal gallery.

Go to Midjourney.com.
Browse to the Explore page or open Your Gallery.
Right-click on any image.
Select “Describe” from the menu.

Once you do this, you’ll be taken to the Create page, and you’ll see four prompt suggestions.

💡 Tip: Look for repeated keywords in the different suggestions. Midjourney often highlights the most important or descriptive words this way, giving you insight into what makes the image tick.

You can:

Click on any of the suggestions to edit or submit.
Submit all four prompts at once.
Modify the wording to suit your style before generating.

Drag and Drop an External Image

If you have an image saved outside MidJourney (maybe from your device or another source), this is the method you’ll want to use.

Follow these steps:

Visit the Create page on the Midjourney site.
Locate the Imagine bar.
Drag your image from your computer and drop it into the bar.

Midjourney will then analyze the image and return four new prompts.

You can even rerun Describe on the same image:

Simply drag and drop the same image again.
You’ll receive a completely new set of prompts.

This helps when you’re not happy with the first round of suggestions or want variety.

Using MidJourney Describe on Discord

If you’re using MidJourney on Discord, there’s a simple way to activate the Describe function directly in your chat window.

Type /describe in the chat.
Choose between:
- Uploading an image from your device.
- Pasting an image URL from the web.
Submit the command.

After a few moments, you’ll get four prompt suggestions listed in a numbered format.

Click the number to submit a particular prompt.
Before the image is generated, you’ll have a chance to edit the text.
Click “Imagine All” if you want to generate all four at once.
Use the Reroll button to get a new batch of suggestions for the same image.

✅ This is helpful if you’re experimenting and want multiple takes on the same image input.

Using Describe in the Full Image Editor

There’s also a third way to access the Describe feature—through MidJourney’s full image editor. This section is for those who enjoy tweaking and experimenting in more detail.

The image editor lets you:

Upload external images.
Retexture parts of an image.
Erase or fill areas based on prompt input.

How to Access Describe Here:

Open the Full Image Editor from the Midjourney website.
Upload the image you want to work with.
Click the “Suggest Prompt” button.

You’ll receive one prompt suggestion only, which appears in the prompt bar.

Once you’ve got the suggestion:

You can edit the prompt before generating.

Use the editor tools to erase, zoom out, or change aspect ratio.

Midjourney will use your edited prompt and existing image data to fill in any empty spaces.

This is great for refining visuals when you have a specific direction in mind.

Key Differences Between the Describe Methods

Here’s a comparison of how each method works:

Feature	Website (Right-click)	Website (Drag/Drop)	Discord	Full Editor
Number of prompts	4	4	4	1
Modify before submit	Yes	Yes	Yes	Yes
Get new suggestions	Yes (repeat drag)	Yes	Yes (reroll)	No
Works with external image	No	Yes	Yes	Yes

When Should You Use Describe?

There are many creative scenarios where Describe comes in handy:

Stuck for words: You’re looking at an image and can’t figure out how to describe it.
Style study: You want to see how Midjourney interprets the style of an image.
Prompt practice: You want to get better at writing effective prompts.
Exploring variations: You want to see different ways to spin the same concept.

🔍 I personally use it often to spot which adjectives or styles Midjourney highlights across various outputs.

Tips for Getting the Most Out of Describe

If you’re just starting out or even if you’re experienced, these tips might help:

Compare prompts: Identify repeated words in the four suggestions. These are likely key terms that influence the image’s style.

Edit for clarity: Don’t feel obligated to use the prompt as it is. Make tweaks to focus on specific elements.

Use in remixing: Grab a prompt from Describe and combine it with others you’ve saved for interesting hybrids.

Reroll wisely: Running Describe multiple times on the same image can surface different interpretations.

Final Thoughts on MidJourney Describe

The Describe feature in MidJourney is simple but super useful. It’s a way to reverse-engineer an image into prompt ideas. If you’re exploring styles, learning how Midjourney “thinks,” or just need a kickstart for your creativity, Describe has a place in your workflow.

There are multiple ways to access it: