Combining the images of two (or even more) in Midjourney is now a fascinating artistic technique for designers, artists and even hobbyists. This technique lets you combine aesthetics, visual content, and “vibe” from two or more pictures into a fresh artwork.
In this article, we’ll look at how to combine two photos in Midjourney step-by-step. We’ll share some of our best practices, common mistakes, as well as a thorough FAQ section that will help you begin with confidence.
What’s the Blend function of Midjourney?
It is possible to use the “/blend” command within Midjourney, which permits users to upload two to five photos and then blend them into one new image.
Instead of the standard image-URL method,/imagine program, /blend is a specific interface to combine multiple images.
Important points:
- Upload 2-5 images.
- By default, the image has an aspect ratio of square (1:1).
- It combines the “concepts and impressions” that are in the photos instead of placing them on top of each other.
- If you are planning to incorporate text prompts with pictures, then you might require the /imagine feature using image URLs (or alternative solutions), as the /blend is only for uploads.
Why should you use image blend?
Image-blending can open up a variety of imaginative possibilities:
- Combining the style of one image and the topic of another
- Merge two photos to generate a novel concept
- Develop hybrid art (e.g. an image + Landscape, two types of illumination, or two art styles)
- Explore possibilities and come up with unexpected results
It’s more about creativity and less about exploring. One writer observes that the /blend “doesn’t provide many useful applications” for extremely controlled work; however, it is excellent for visual exploration and inspiration.
How to Mix Two Images in the middle of a journey: Step-by-Step Guide
Here’s a complete walkthrough.
1. Make sure you have your photos ready
- Select two images you wish to mix (you can mix as many as five).
- For greater consistency, choose images that have similar aspect ratios (e.g. the same size, either portrait or square). This will help avoid awkward framing and empty areas.
- Check that the images are of acceptable quality (good resolution, clear subject) to ensure that Midjourney has enough information to work with.
2. Open Midjourney bot/channel
- If you are on your Discord server or using the Midjourney bot, ensure that you are in a channel that the bot has access to.
- Check that you have your Midjourney subscription current and that you have access to the bot.
3. Use the command /blend
- Within the input text field (in Discord), enter: /blend.
- When you log in, you’ll be presented with an option to upload your images (drag-and-drop or choose).
- Upload images one as well as picture 2. (minimum of two). It is possible to add additional photos (up to 5 images in total).
- You can choose the dimensions (square or portrait Landscape). It’s by default rectangular (1:1).
4. Send the command, and wait for the generation
- Press Enter or Send after uploading the photos. The bot will start processing, and after a while, will provide a set of results (usually four variations).
- Examine the results. If you are satisfied with one, you can increase or alter it more; if not, try rerunning with modifications.
5. Refine and increase
- After a decision has been made and you are happy with it, you can make use of the Upscale (U) as well as Variations (V) to fine-tune.
- If you’d like to add a prompt with text or control style, a good option is to do so after the initial blend, switch to the “Remix” option or change to a prompt which incorporates the result of the image. Specific workflows incorporate the /imagine feature using image URLs and prompts with text to provide more control.
Tips and Best Practices
To make the most of the blend, there are a few practical suggestions:
- Find aspect ratios that match: If the image is a tall image and picture 2 has a vast landscape, the results could have odd framing. Utilise similar ratios or define the dimensions in advance.
- Image quality is crucial—clear, high resolution and minimal cropping yield more results.
- Make use of more than one style: If you’d like one style to be dominant, you can upload the image twice (if permitted) or even combine three images, where two are representing the style, and one is the subject. Some users have reported that this adds more weight to the image of style.
- Try out prompts following the blend: Since the blend doesn’t support prompts with text during upload, it is recommended to end by prompting or mixing. One solution is to utilise /imagine with images and prompts when you require greater guidance.
- Expect uncertainty: Blends may surprise you. Sometimes, the result diverges from what you’d expected. This is part of the attraction.
- Remix and repeat: Utilise “Vary” or other prompts to improve compositions, lighting, or other aspects.
- Make use of fewer images for clarity: If you upload a lot of images (e.g. five images or more), the output may be confusing. Three or two images usually produce better, clearer output.
Common Errors to Avoid
- Uploading images in drastically different formats or aspect ratios can result in unnatural compositions.
- You must be sure that you have control over the elements of each image that take the lead. Midjourney blends images in a variety of unpredictable ways. According to one user: “You can combine any two images using ‘blend’. … But … one thing that you aren’t able to decide is which features from an uploaded image(s) MJ incorporates into the final.”
- Try to include an option to type in a text message while you are using the /blend upload. Text isn’t natively supported within the upload interface. You might need to change it to a different method.
- Inadvertently setting dimensions when you want to use a non-square ratio.
- Uploading poor-quality or heavily edited images can result in less desirable outcomes.
A Use Case Example
Imagine you have:
- A photograph of a city’s skyline at night
- A painting depicting an eerie dreamscape
You can upload both photos through Blend, then set the size in Landscape (3:2), and then upload. Midjourney will blend the architectural elements of the skyline and the surreal colour palette, or the texture that creates the dreamscape. After the creation, you can select a different version, scale it, and then remix it with the prompt, such as “cinematic lighting or volumetric fog, super detail”. This workflow lets you combine art and realism all in one.
Final Thoughts
Blending images in Midjourney is a powerful creative tool that sits somewhere between art exploration and design experimentation. When done thoughtfully, with attention to image quality, aspect ratios, and desired stylistic outcome, you can achieve visually compelling results that push beyond what simple prompt-based generation offers. At the same time, it’s essential to approach /blend with curiosity and an experimental mindset: enjoy the surprises it delivers, then refine from there.
If you’d like, I can also walk you through advanced techniques (e.g., image-weighting, mixing more than two images, combining blended output with prompts) or show real-world examples of successful blends.
FAQ (Frequently asked questions)
1. What is the maximum number of images I can combine at the same time?
It is possible to upload a maximum of 5 images using the /blend command.
If you’d like to make use of more than five images, you’ll have to employ the approach of the /image method with multiple URLs for images.
2. Do I have to add an option to add a message along with the images blended?
Utilising the standard blend interface, it is not possible to add a text prompt in the same manner, as it is based solely on uploads of images.
However, you can:
- Make use of the word “imagine” by combining image URLs with prompt text
- After mixing, you can use the Remix or Vary function to incorporate prompt text to refine the result further.
3. Which aspect ratios can be available?
The default result will be rectangular (1:1). It is possible to select the portrait (2:3) as well as the Landscape (3:2) by choosing the Dimensions setting. It is recommended that you upload images that are in line with your desired aspect ratio in order to avoid accidental cropping or framing problems.
4. What does Midjourney use the process to decide what blends to mix?
The AI analyses the pictures’ “concepts and vibes” and combines their style, compositions and colour schemes. The input images influence the result. It is not possible to explicitly say “use the image’s shape one and colour from image 2” without a guarantee that the results are primarily exploratory.
5. Blending can be helpful in professional tasks?
It is possible to. Although the term “mix” is typically regarded as “fun and experimenting”, it also has potential for conceptual art, visual conception, character design, mood boards, as well as hybrid visualisations. However, when you need precise visuals or a brand-aligned composition, you may require additional development in other tools after the initial generation.
6. What is the situation with image rights as well as licensing?
Make sure you’re authorised to make use of the photos you upload (especially in the case of copyrighted images). The resulting blend could retain the distinctive features of those inputs, and therefore, the responsibility of the user is a must. Check out Midjourney’s terms and conditions of service applicable to commercial use.
7. Do I have to combine multiple images, and is there a drawback?
Yes, up to five images can be posted. However, more images could cause more chaos as the “blend” will become more of a mix, but less of a precise blend. One guideline states that blending three or four images may still be effective; however, it may need more work for clarity.
8. What happens if the pictures are different (e.g. one is an illustration, while the other is a photograph)?
It is possible to mix a variety of images. The result might appear more abstract or surreal, and that’s what you’re looking for. However, if you’re looking for an even-handed result, using images that share the same visual language (lighting composition, light and subjects) is often helpful.