What to Expect in Midjourney V7 Alpha: Your First Look

Midjourney version 7 Alpha is here! It's an early look at what's coming next. This release is still in testing, so some parts work great, and others... not so much. Let's dive into what you'll find and what features are still being built out over the coming weeks and months.

Setting Up Your V7 Experience: The Personalization Profile

When you first switch on V7, Midjourney asks you to train a personalization profile. You have to do this; you can't skip it. You need to rate at least 200 image pairs.

How you do this depends on whether you care about having a specific personalized style.

• If you just want to get through it quickly, just click randomly on image one or two.

• If you want to set up a profile that matches your taste, look closely and select the left or right image that you like better.

• If you hate both images in a pair, you can skip it by hitting three.

Don't worry too much about getting it perfect the first time. You can create another profile later if you want.

Meet the New Profile Menu

After you finish the training, you'll see a change in the interface. There's a new "P" button on the create bar. Clicking this brings up a handy menu. This lets you turn your profiles on and off, switch between different profiles you've made, and even add mood boards.

You can use a personalization profile and mood boards at the same time, or even multiple mood boards together. This menu is a great addition for managing your styles.

Features Still Getting Ready in V7 Alpha

Remember, V7 is alpha. This means many features from V6.1 aren't fully working in V7 yet. They are still ironing things out.

Moodboards and Style References (SREF)

Two features many people use for style are mood boards and SREF images. In V7 Alpha, you can still use these. There's a new way to add SREF images now – just drag them into the style reference spot in the create bar.

However, the results aren't great right now. They seem to pull less style information than they do in V6.1. The team says they are still working on making these better in V7.

A Quick Note on Switching Models & Personalization

There's a bug to watch out for. If you use personalization with V7 Alpha and then switch back to the V6.1 model, personalization might stay on even if the "P" button is off. To fix this, type `dash dash P none` (`--p none`) in your prompt. This forces personalization off.

Managing your prompts and settings manually can be time-consuming. If you're generating many images and need to manage different styles, models, and settings, consider exploring tools that can automate parts of your workflow. The Midjourney Automation Suite from TitanXT helps streamline your image generation process.

Experiencing the New Speed Modes

Speed modes are a bit different in V7 Alpha right now.

• Relaxed mode is available.

• Turbo mode is available.

• Fast mode is not working as a distinct speed mode; it's currently the same as Turbo.

A new Fast mode is coming soon, but for now, let's talk about the exciting new Draft mode.

Discovering Draft Mode

Draft mode is a brand new way to work. When you use `dash draft` with the Turbo setting selected (remember, Fast is currently Turbo in V7 Alpha), Midjourney generates images incredibly fast. It feels like almost real-time generation. It's impressive how quickly it produces results.

The speed comes with a tradeoff: the image quality isn't high. You can't currently upscale or rerun a specific image generated in this mode to improve its quality later. However, Draft mode is very useful for quickly testing prompt ideas. It costs only a fourth of the credits of a standard generation.

A handy trick: once you find an image you like from a Draft mode batch, go back to the standard V6.1 model and use the `dash dash subtle` parameter with the same prompt. This often creates a similar image in better quality.

The Draft Mode Button: Conversational AI

To the right of the create bar, you'll see a "Draft Mode" button. Turn this on, and the button turns red. This activates a conversational mode. You can now talk to Midjourney about the last set of images you created.

You can copy your original prompt and run it. Then, you can give Midjourney instructions like "make this a one:1 aspect ratio," "keep the colors but make it a photo," "zoom in closer," or "make it a woman and make her happy." As you type these requests, Midjourney will adjust the prompt and generate new images based on your conversation. There's even a microphone icon if you prefer to speak your instructions.

Once you have the prompt perfected through this conversation, you can turn off Draft mode and refine the prompt using the standard settings, or use the "enhance" button (which effectively does the same thing).

While you can't enhance a specific image from the conversational draft session right now, this mode is fantastic for rapid prompt experimentation and refinement.

Refining prompts manually can be tedious. Imagine having a system that helps you manage these iterative processes or even automate the variations you explore. The Midjourney Automation Suite by TitanXT offers powerful ways to accelerate your workflow and generate more creative options efficiently.

Exploring the Core V7 Model Quality

Despite the features that are still in progress, the base V7 model shows great promise. Just by prompting, you can see the quality that's coming.

It seems to produce better emotional depth in images compared to V6. It also handles coherence well, meaning it gets more elements right when you ask for specific things like colors, subjects, and their actions within a scene.

The level of detail is impressive. Images generated with the V7 model look sharp and have a new elegance. The output quality of the images themselves is a strong point of this alpha release.

What Else Is Coming Later?

[P]Several important features from V6.1 aren't fully functional in V7 Alpha yet because they use V6 technology. They are waiting for V7 versions.[/UL]

[LI]Image Editor (Panning): Currently uses the V6 engine, which means you lose V7 quality when expanding images.[/LI]

[LI]Upscaler: Still uses the old upscaler. A new, potentially faster, cheaper, and higher-resolution (maybe 4K) upscaler is expected.[/LI]

[LI]Character Consistency: This important feature will be part of a new system called "Omni Reference," which is Midjourney's flagship function for V7. It's not available yet.[/LI]

[/UL]

New versions of the editor, upscaler, and the Omni Reference system are expected in the coming weeks or months.

Is V7 Alpha Ready for Your Projects?

For serious work where you need full functionality and predictable results, it's probably best to stick with the V6.1 model for now. It's fully functional and stable.

However, V7 Alpha is definitely worth playing with for fun and to see the quality of the base model. Check the explore page; people are already creating some amazing things with it.

The outlook for V7 is exciting. Many new features are coming, and the core model quality is already very high. 2025 looks promising for Midjourney users.

If you're looking for ways to manage your growing library of Midjourney creations, test prompts efficiently, or automate repetitive tasks, consider how automation can help. The Midjourney Automation Suite by TitanXT can help you streamline your workflow and unlock new possibilities with your images.