Gallery

Contacts

405 W. Greenlawn Ave Lansing, Michigan 48910

contact@techjacksolutions.com

+1-616-320-4064

Midjourney

Midjourney V8.1 Explained: What the Latest Version Adds

If you searched for Midjourney V7, here is the first thing worth knowing: V7 is no longer the latest model. The current stable version is V8.1, released in alpha on April 14, 2026. It arrived less than a month after V8 shipped on March 17, and it builds directly on the aesthetic direction V7 introduced back in April 2025. So the V7 you may have read about is part of the story, not the end of it.

This breakdown walks the V7 to V8 to V8.1 progression honestly, then explains the four things V8.1 actually adds: HD images at roughly 2K resolution with no upscale step, restored Image Prompts, a Prompt Shortener, and an updated Describe tool. It also covers the one setup step that catches people off guard, the Global V7/V8 Personalization Profile you have to turn on first. The goal is simple: by the end you should know which version you are on and why it matters.


V8.1
Current Stable
Apr 14 2026
V8.1 Alpha Release
~2K
HD Resolution via --hd
2022
Open Beta (July 12)

What Is the Current Version of Midjourney?

The current stable version is V8.1, which Midjourney released in alpha on April 14, 2026. It succeeded V8 (March 17, 2026), which in turn replaced V7 (April 4, 2025). When you read older guides that treat V7 as the newest model, they are simply dated. The model picker in both the Discord bot and the web app at midjourney.com now offers V8.1, and that is where new image quality work is happening.

It helps to know what Midjourney is before getting into version differences. It is a subscription image generation service from Midjourney, Inc., a small San Francisco research lab founded by David Holz, which opened to the public in beta on July 12, 2022. You write a natural-language prompt, and the model returns images. Since 2025 it also generates video from images. There are two ways in: the Discord bot, where /imagine plus a prompt returns a grid of four images you can upscale or vary, and the web app, which adds an editor for pan, zoom, Vary Region, and inpainting. The two stay in sync.

Practitioner note: The version number is not cosmetic. Each major release changes how prompts are interpreted, so a prompt tuned for V6 can read differently on V8.1. When you upgrade, expect to re-test your go-to prompts rather than assume they carry over one to one.


The Midjourney Version Timeline

Tracing the releases makes it obvious where V8.1 sits and why V7, while still selectable, is no longer the front edge. A few milestones matter more than others: V4 (November 2022) was the point where Midjourney moved its model training onto Google TPUs, V6.1 (July 2024) launched the web UI, and the V7 through V8.1 run is the recent acceleration.

Feb to Jul 2022
V1 to V3, then open beta
V1 (Feb), V2 (Apr), and V3 (Jul) arrived in quick succession, with public open beta opening July 12, 2022.
Nov 2022
V4: training moves to Google TPUs
From V4 onward the models are trained on Google TPUs. This is the release where output quality took a visible step up.
Mar 2023 to Dec 2023
V5 line and V6
V5 (Mar), V5.1 (May), and V5.2 (Jun) refined coherence and added editing such as Vary Region. V6 landed in December 2023.
Jul 2024
V6.1: web UI launches
V6.1 shipped alongside the web app, ending the Discord-only era and adding the browser editor.
Apr 4, 2025
V7: the aesthetic still in use
V7 set the look that V8.1 carries forward. It is still selectable but is superseded as the default direction.
Mar 17, 2026
V8
V8 succeeded V7 and set the foundation that V8.1 extended weeks later.
Apr 14, 2026 (current)
V8.1: the latest version
The current stable release. Adds HD images, Image Prompts, a Prompt Shortener, and an updated Describe tool on top of the V7/V8 aesthetic.

HD Images: 2K Without the Upscale Step

The headline addition in V8.1 is HD image generation. By adding the --hd parameter to a prompt, the model produces images at roughly 2K resolution directly, with no separate upscale pass required. In earlier workflows, getting a higher-resolution result meant generating a base image and then running an upscaler on the one you liked. V8.1 folds that into the initial generation.

For practitioners, the change is about iteration speed and predictability. When the upscale is a second step, you sometimes find that the upscaled image drifts from what you approved at low resolution. Generating at HD from the start means the composition you see is closer to the final composition you keep. The tradeoff is that higher-resolution generation is more computationally expensive, so reach for --hd on the prompts you intend to finish, not on every exploratory draft.

--hd
Add this parameter in V8.1 to generate at roughly 2K resolution directly. No separate upscale step. Use it on prompts you mean to finish, since HD generation costs more compute.

Resolution is not the only quality lever in V8.1, but it is the most concrete change for anyone who delivers finished assets. If your prior habit was always running an upscale at the end, V8.1 lets you bake that intent into the prompt instead.


The Three New Tools in V8.1

Beyond HD output, V8.1 ships three workflow tools. None of them reinvents how you prompt, but each removes friction from a task that practitioners do constantly.

Image Prompts
Feed an image as part of the prompt to steer composition and style
Input Image + text
Status in V8.1 Supported
Use for Visual anchoring
Prompt Shortener
Trims a long, overloaded prompt down to what actually drives the output
Input Verbose prompt
Output Tighter prompt
Use for Cleanup
Updated Describe
Reverse-engineers candidate prompts from an uploaded image
Input Uploaded image
Output Prompt text
Use for Reverse prompting

The Prompt Shortener is the one most likely to change a daily habit. Many practitioners build up sprawling prompts over an editing session, and the long tail of descriptors stops earning its keep. Shortening the prompt makes results more reproducible and easier to reason about. The updated Describe tool, meanwhile, is useful when you have a reference image and want a starting point in Midjourney's own prompt vocabulary rather than guessing at it.


The One Setup Step: Personalization Profile

Here is the part that trips people up. To use V8.1 you must first turn on the Global V7/V8 Personalization Profile. This is a one-time ranking step: Midjourney shows you pairs of sample images and you pick which you prefer, building a taste profile the model then applies to your generations. Until that profile is active, the version cannot apply its personalized behavior.

If you are coming from an older version where personalization was optional, this can feel like an extra hurdle. In practice it takes a few minutes, and it is the mechanism that lets V8.1 lean into your preferences rather than a generic default. The profile is shared across V7 and V8, so setting it up once covers both model families.

Profile must be active before V8.1 personalizes output
If your V8.1 results feel generic, check whether the Global V7/V8 Personalization Profile is actually turned on. The ranking step is easy to skip on first run.
Alpha release, so expect change
V8.1 shipped as an alpha. Behavior and parameters can shift as it matures, so confirm current details on Midjourney's own documentation before relying on a specific flag.

Where Does V7 Fit Now?

V7 is not dead, and the search interest behind the term is real, but it is no longer the latest version. Its main role today is as the aesthetic ancestor of V8.1: the look V7 introduced in April 2025 is the look V8.1 refines. If you spent time learning V7's prompting behavior, that knowledge transfers, because V8.1 sits in the same lineage rather than starting over.

So when someone asks "should I use V7 or V8.1," the honest answer for most workflows is V8.1. You keep the V7 sensibility you may already be comfortable with, and you gain HD output, the restored Image Prompts, the Prompt Shortener, and the updated Describe tool. The reasons to deliberately pin V7 are narrow, mostly reproducing a specific older result. For new work, V8.1 is the default.

Practitioner note: If a tutorial or prompt pack is labeled for V7, you can usually run it on V8.1 with minor adjustments rather than treating it as obsolete. Test, then tune. The aesthetic continuity is the whole point of the V7-to-V8.1 line.


Should You Switch to V8.1?

For nearly everyone generating new images, yes. The HD output alone removes a step from most finishing workflows, and the three tools reduce friction without forcing you to relearn prompting. Set aside a few minutes to turn on the Global V7/V8 Personalization Profile, re-test your favorite prompts, and decide where you want --hd in your routine.

Treat V8.1's alpha status as a reminder to keep one eye on the source. Because versions and parameters move quickly, the safest habit is to confirm the current behavior on Midjourney's own pages before you commit a flag to a production pipeline or a client deliverable. If you are still deciding whether Midjourney fits your work at all, start with our broader explainer of what Midjourney is, then come back to the version specifics here.


Frequently Asked Questions

What is the latest version of Midjourney?

V8.1, released in alpha on April 14, 2026. It followed V8 (March 17, 2026), which succeeded V7 (April 4, 2025). V7 is no longer the latest version, even though the keyword still draws search traffic.

What does V8.1 add over V8 and V7?

HD images at roughly 2K resolution generated directly through the --hd parameter with no separate upscale step, support for Image Prompts, a Prompt Shortener that trims verbose prompts, and an updated Describe tool. To use V8.1 you must first turn on the Global V7/V8 Personalization Profile.

Is Midjourney V7 still usable?

Yes. The V7 model remains selectable, but it is superseded. V8.1 inherits the V7 aesthetic lineage while adding HD output and the new tools, so most users should move to V8.1 for the resolution and feature gains.

Do I need to do anything special to enable V8.1?

Yes. V8.1 requires turning on the Global V7/V8 Personalization Profile, a one-time ranking step where you rate sample images so the model can apply your taste profile. Until it is active, the version cannot apply its personalized behavior.

What hardware are Midjourney models trained on?

From V4 (November 2022) onward, Midjourney models are trained on Google TPUs.

Fact-checked against vendor documentation and official sources, June 2026
Midjourney is a trademark of Midjourney, Inc. Discord is a trademark of Discord Inc. Google and TPU are trademarks of Google LLC. Gemini is a trademark of Google. All other trademarks belong to their respective owners.
Before You Use AI
Your Privacy

Midjourney is a subscription service, and the images and prompts you submit are processed on its infrastructure. Prompts, uploaded reference images, and generated results may be retained and, on some tiers, visible to other users in shared galleries unless you are on a plan with a private generation mode. Review Midjourney's terms and privacy documentation before submitting sensitive or proprietary imagery, and confirm what is public on your specific plan.

Mental Health & AI Dependency

Generative image tools can blur the line between reference and reality, and heavy reliance on AI for creative output can erode your own judgment about quality and originality. Keep a human in the loop on anything you publish or sell. If you or someone you know is experiencing a mental health crisis:

  • 988 Suicide & Crisis Lifeline -- Call or text 988 (US)
  • SAMHSA Helpline -- 1-800-662-4357
  • Crisis Text Line -- Text HOME to 741741

AI systems can produce plausible-sounding but incorrect guidance. For mental health, medical, legal, or financial decisions, always consult a qualified professional.

Your Rights & Our Transparency

Under GDPR and CCPA, you have the right to access, correct, and delete personal data held by a service provider. Midjourney's training data and image rights have been the subject of ongoing copyright litigation, so review the current terms before commercial use. Tech Jacks Solutions maintains editorial independence. This article was not sponsored, reviewed, or approved by Midjourney, Inc. or any vendor mentioned, and we receive no affiliate commissions from any linked provider. The EU AI Act sets transparency expectations for generative systems. Our evaluations are based on primary documentation and verified data.