Latest Midjourney Statistics: Growth, Usage, and Creative Trends

The economics of production are imploding (and this is what that means)

Let’s throw some rough figures on that. When you commission a professional designer to create an illustration for you, you’re likely to pay anywhere from $100 to $1000+ (depending on the intricacy of the design). With Midjourney, you pay a monthly subscription fee, usually between $10 and $60, and create tens or even hundreds of images a month.

This is not just a reduction in price. This is a fundamentally different business model.

According to Business of Apps, Midjourney is already making over $200 million a year in subscription fees. This illustrates just how powerful the low-cost, high-volume business model can be.

An odd equilibrium. Increased competition, but more opportunities as well, even if not necessarily in the exact same way as in the past.

The artist’s position is evolving, not going extinct

I keep seeing the idea floating around that AI will completely take over the jobs of artists. That I don’t believe. I do believe, however, that the value proposition of artists will change.

How to execute used to be the big question. Now? Taste, direction, and originality are more valuable. Anyone can create an image. Not anyone can create a meaningful one.

Indeed, professionals are increasingly using Midjourney as a tool, not as a replacement, for accelerating idea development, testing concepts, testing visual styles.

Research shows a steady increase in the number of designers incorporating AI into their design processes, not replacing their design processes, (according to Statista.

Honestly, I think that’s a much more constructive way to approach the matter. Not “AI or artists.” But “AI and artists.”\

Pricing models are being rewritten in real time

One more silent change, pricing of creative services.

If clients know you can draft things rapidly with AI, they will expect more efficient turnarounds… and sometimes lower prices. That’s the scary part.

However there is an upside. Creators who can leverage AI tools can provide even more value, more rapid iterations, wider exploration, and hybrid processes.

It’s like this:

It’s not simple. It takes adjusting, it takes learning new things and sometimes it takes scrapping the way you have done things for years. But what’s the real risk? Ignoring it?

So what is art “worth” now?

That’s the question that seems to keep coming up and my answer to that is… it depends. Are you determining the value of it based on the amount of time or effort it takes to create it? That seems to be where AI will upset the apple cart.

Are you determining the value of something based on how it makes you feel, the originality of it or the message it conveys? Then I don’t think AI has changed anything. That picture that was made with Midjourney may be amazing… but does it have a message?

Does it evoke emotion in people who look at it? Is it original or is it one of 1000 pieces that look just like it? That’s where I think the real value of human artists still lie. Perhaps that’s the point. I don’t think AI has made art worthless, I think it has made us reevaluate what it is that we value about it.

Global Usage Breakdown: Where Midjourney Is Growing Fastest (By Country & Industry)

Global Adoption: The Countries and Industries Most Embracing Midjourney

Beyond Silicon Valley

There was a time when you could safely assume that Midjourney was a Silicon Valley phenomenon. That’s where all the tech originates and then disperses, right? Well, not so much these days.

Adoption of Midjourney has become more global. Not just in some nebulous sense that “there’s lots of international users” but in the more concrete sense that there are whole countries that have grabbed onto it and started running.

More than 60% of Midjourney users are outside North America, according to DemandSage.

But you could already tell that based on some of the prompts and styles and cultural influences. It’s less dominated by Silicon Valley than it used to be, and more, well, distributed.

Countries most engaged with Midjourney

Here is a country by country look at Midjourney use. The base is fairly wide, though there are some countries that are growing faster than others.

APAC region

APAC is really interesting. India, Indonesia, South Korea, etc. are all growing really fast. Partially because they skew a bit younger and are more digitally active. Europe is 2nd, led by creative and freelance communities.

North America is a big chunk, but we’re no longer the epicenter. The center of gravity is moving. See Statista’s larger AI adoption data here.

Fastest growing countries

Not what you’d expect. Obviously, you’d think that the US, UK, Germany, etc. are all growing really fast. That is true. But there are a number of countries that aren’t usually in the headlines that are growing even faster. For example:

• India
• Brazil
• Philippines
• Turkey

These countries are all growing really fast, particularly among young creatives and freelancers. Why? Lower friction. You don’t need a fast computer or go to school for two years. You just need a web browser and a curiosity.

“Emerging markets are adopting generative-AI applications faster than some developed markets, driven by the greater availability of these tools and the stronger need for additional digital-income streams,” according to a report by consulting firm McKinsey. I dunno. That’s sort of cool to me. Less gatekeeping in the creative world.

Industry breakdown: Who’s actually using it at scale

Now here’s where things get more practical. Which industries are actually using Midjourney regularly-not just experimenting, but integrating it into workflows?

Marketing is the obvious frontrunner here. There’s a never-ending appetite for visual content. Social media alone is hungry for images. Over 40% of marketers now use AI for visual content creation, as per HubSpot.

As soon as something saves you time (and money) it becomes the new normal. No-one is going to go back to something that is slower than that.

Freelancers and solo creators: The real growth engine

This is where I don’t see enough being said. Businesses are using Midjourney of course, but a lot of growth is coming from individuals. Freelancers, side hustlers, content creators. People who don’t have budgets, people who don’t have a team. Midjourney helps.

Now one person can do the work of a designer, an illustrator and a concept artist. The majority of AI tool users are independent or small-scale creators, as per Coolest Gadgets.

And if you have ever had to try and do all those things by yourself, you’ll appreciate how valuable that is. It’s not just a time-saver, it’s an enabler.

Localised Cultures: Style is dependent on the region of the world you are in

It is underappreciated the fact that culture and geography also play a massive part in what people are creating. It isn’t just the spread of midjourney that is happening, it is diverging in different regions of the world:

• Asia: Anime, videogames and stylised art
• Europe: Arty, experimental, surreal
• North America: Commercial, polished, advertorials
• LatAm: Colourful, lively, culturally strong

This is part of the reason why I find the platform fascinating. It isn’t just that there is more of everything, it is that you are seeing more varied ideas. Sometimes they collide and combine in weird ways, but that is the nature of the beast. It is human and unpredictable.

Where is Midjourney gaining the most traction then?

The simple answer is: emerging markets and the creative fields. Regions where the opportunity to make a living is vast, where digital competencies are the fastest route to a paycheck and where the means to create are not restricted by the ability to afford expensive tools.

That’s where Midjourney is making the biggest inroads.

Commercially, the early adopters are in time-sensitive fields such as marketing, content, and commerce. In a broader sense, the creative economy is globalizing, decentralizing, and, well, becoming more democratized. And that seems to be more significant than one tool.

Most popular Midjourney styles, keywords, and creative trends

Prompt Power

It always starts with a bad prompt. We’re just going to get that out of the way. No one writes the perfect prompt first try. You type in something like “cool futuristic city” and hope for the best… and what you get is either genius… or hot garbage. Sometimes both.

Then you tweak it. Add “cinematic lighting.” Maybe “ultra realistic.” Suddenly it’s better. You feel like you’ve unlocked some kind of cheat code. Truth is, Midjourney is just a prompt game. And the better and better people get at it, the more trends you start to see.

According to various AI usage reports, people are using multi-keyword prompts with style modifiers more than simple descriptions, as is clear when you compare Midjourney vs Dalle vs Stable Diffusion.

The most popular styles (and why they won’t go away)

Take a look through Midjourney galleries for five minutes and you’ll start to see the same styles repeated over and over again. Not because people aren’t creative, but because some styles just slap.

Here are some of the most popular styles:

Hyperrealism is king. This is the most obvious of all the trends. The truth is, humans are suckers for photorealistic images. It’s a really simple trick to get people to get excited about an image. “Whoa, look at that! It’s so photorealistic! That’s so cool!” Hyperrealism will always get a higher wow factor than anything else.

Stylized visuals, anime, cartoon, all that stuff is still huge. Especially in Asia, especially among younger generations, as Statista reported in their digital culture survey. . Anime, cartoon and especially cyberpunk are all still huge, but I’m just giving you the most popular stuff.

Hyperrealism is the most popular. We talked about anime being huge and there’s a lot of overlap between the different trends. “Hyperrealistic anime cyberpunk portrait” is a thing and it works.

Overused keywords

Now I’m going to share with you some keywords that are used so frequently that they are now basically part of the Midjourney language. I’m sure you’ve probably used many of these without even realizing it. Keywords like:

• cinematic lighting
• ultra detailed
• 8k
• photorealistic
• sharp focus
• dramatic shadows
• depth of field

These are not accidents. These are tools that help push the AI in a certain direction. Data suggests that prompt complexity and length have increased significantly over time, with users stacking multiple modifiers to refine outputs, according to EarthWeb.

It’s kinda ironic, really. We had to develop a whole language just to communicate with a robot.

Trends come and go… fast

One week it’s “Wes Anderson style.” Next week it’s “AI-generated fashion models.” Then suddenly everyone’s doing “ancient statues in modern settings.” Trends in the Midjourney community move at an insane pace. Even faster than in the rest of the art world.

Because iteration happens in real-time. Social media is a big part of this. When one popular post goes viral, it creates thousands of copy-cat prompts in a matter of hours.

According to HubSpot’s social media trend analysis, AI-generated visual trends spread across social media platforms in a matter of days not months. Sometimes it gets boring. But then something new comes along and changes the game all over again.

The rise of “prompt engineering” (yes, it’s a thing now)

This is something that still feels a little awkward to say aloud: writing prompts has become a skill.

People share prompt templates, sell prompt packs, even offer courses. And before you roll your eyes, it actually makes sense.

The difference between a basic prompt and a well-crafted one can be massive.

Some artists are able to regularly generate amazing images, not because the tool is different, but because they know how to communicate with it.

Now this, I think, is where it gets a bit more human again. The role of skill hasn’t vanished. It has simply changed.

What does that tell us about creativity?

From a distance, I think it is quite remarkable. We are seeing the birth of a new language in action. Not painting, not coding, but something inbetween.

We’re not just generating images, we are learning to speak in ideas, more accurately, more visually, more emotionally.

And yes, sometimes it feels a bit mechanical. Trial and error, repeating words, fiddling with phrases. But other times, it feels like working together. Guiding the machine.

If you ever spent 15 minutes editing a prompt just to get the lighting right, you know what I mean.

Perhaps the trend isn’t a style or a keyword at all. Perhaps the trend is simply this transition. From designing with your hands to designing with your mind first. And honestly, that’s no minor thing.

Inside the Algorithm: How Midjourney Generates Images and Why It Keeps Improving

The Magic in the Machine

It’s a mind-blowing feeling… until you have to describe it.

You input a prompt. Then you wait a few seconds and an image is returned that resembles something a professional photography studio would produce. The initial thought is: this is insane, how does this work?

The short answer is that it’s not magic. The longer answer is a bit more complicated. MJ uses something called a “diffusion model”. A diffusion model starts with pure noise and iteratively refines it until it creates an image based on the prompt you input. Think of it as chiseling a sculpture out of white noise.

The model itself is trained on a dataset of image-text pairs, so it understands the relationship between certain words and visuals. When you input “a cinematic portrait with soft lighting” the model isn’t just making something up, it’s based on its understanding of a million (or a billion) examples for a summary on generative AI models.)

Noise to Image: A Step-By-Step Explanation

Here is a more simplified explanation that doesn’t require a Doctorate:

• You input text
• The model interprets the text into a mathematical equation
• It begins with noise
• It slowly denoises and refines the image
• The process is repeated, and eventually, you have an output.

This happens nearly instantaneously…often in seconds. It’s a bit ridiculous when you think about it. Complex math, probability, and pattern recognition…distilled into a progress bar. And yeah, it still feels a bit like cheating sometimes.

Why the results keep getting better

Now this is the part that’s really interesting. The first few versions of Midjourney were great…but you could tell they were AI generated. The faces weren’t quite right, the hands were…well we don’t even talk about the hands.

Now? You can see the difference. The images are sharper, the anatomy is better, the style is more consistent. So what changed? Three big things:

• Better training data
• Improved model architecture
• User feedback loops

AI models improve as they are trained on more diverse and higher-quality datasets, and as developers refine how the models interpret prompts, according to McKinsey’s analysis of generative AI development.

But there’s also something else…users themselves. Every prompt, every selection, every preference feeds back into the system in some way. It’s millions of tiny corrections happening all the time.

Each new iteration gets closer, with fewer mistakes and a better understanding of the prompt. It isn’t there yet, but it is getting closer very quickly. To be honest, that can be a little disconcerting at times. You get used to a certain level of quality, and then it improves again.

Why prompts are more important than you realise

One thing that people often seem to forget is that it’s not all the algorithm. The prompt matters. If you ask for something vaguely, you’ll get something vague. If you ask for something specific, you’ll get something more specific.

According to EarthWeb, researchers found that the quality of text prompts had a direct effect on the quality of the results, with more detailed text prompts leading to more accurate results that were more visually pleasing.

So when you say “AI did it,” well, that isn’t strictly true. AI assisted. But you told it what to do. Sometimes you have to fight with it, but you told it.

Those imperfections are a feature of the product

I’ve mentioned a few ways that Midjourney v5 improves on v4. But it is still far from perfect. It can’t get text right, or handle multi-part actions, or follow highly specific directions.

Perhaps that’s okay.

The imperfections serve as a reminder that there is still a distance between the imagined and the realized. A distance in which humans can play a role. We can’t lose sight of what the AI can’t do.

So… is the algorithm “creative”?

Regardless of whether we want to, we have to get philosophical here.

Midjourney isn’t imaginative in the way that humans are. It doesn’t have emotions. It doesn’t “desire” to make things.

But it can synthesize concepts in creative ways. Sometimes, that can be enough to make the distinction fuzzy.

Ever seen a generated image and thought, “I didn’t ask for this, but I kinda like it?” You’ve lived the fuzziness.

Perhaps the real question isn’t whether the algorithm is creative, then. Maybe it’s how it alters our own creativity.

Who is using Midjourney? Demographics, Occupations and Other Interesting User Behavior

It’s not just the “tech folks” anymore

When people would have been asked who is going to use Midjourney back in 2022, they would probably have guessed designers, possibly some developers and the odd digital artist. A fair guess… except that that’s so last season. We are in a new season. The user base has long since expanded beyond its original niches.

Midjourney nowadays doesn’t feel like a tool and users feel like a digital art studio that you can visit and in which you can tinker. Estimates tell us that the majority of Midjourney users are non-professionals.

That is, they don’t work as designers or in the creative industries full time. And well… that’s the story. The tool didn’t just scale, it democratized.

Demographics: Age

Young, but not entirely so

You would think this is a Gen Z platform. It is, but not as much as you think.

25 to 34: No surprises here. They were born with the internet and are heavily invested in creative and professional applications. But here’s a fascinating age group. People who didn’t grow up with design tools are now experimenting with visual creation for the first time.

So that’s exciting. Overall AI adoption follows a similar pattern, reaching beyond younger generations. Occupations: Not what you’d think. It’s not all “creative types”.

Professions: Not what you’d expect

This one’s a bit of a twist. The jobs of Midjourney users don’t really fall into a single bucket. It’s a melting pot of creatives and non-creatives.

Hobbyists are actually the biggest group. That still surprises a lot of people. Designers are obvious. But they are not the dominant type of user any more. Marketers and content creators are coming on pretty strong too.

We have a lot of users who use it for marketing campaigns and image for thumbnails and branding ideas. And then there’s just “other” that’s teachers, writers, entrepreneurs, that kind of thing.

People who never identified as visual artists before. It seems that AI software is being used by most professions, particularly those in content, says HubSpot.

The Future of AI Art: Midjourney Predictions for 2030 and Beyond

We’re not even close to the “final version” yet

It’s tempting to look at Midjourney today and think, “Wow, this is already peak.” But that’s probably the same thing people thought a couple of versions ago, and then everything improved again.

The reality is, we’re still early. Like… early early.

Generative AI is expected to contribute trillions of dollars annually to the global economy by 2030, which includes creative industries like design, marketing, and media, according to McKinsey.

That kind of scale doesn’t happen unless the technology keeps evolving. And if it keeps evolving, Midjourney (and tools like it) won’t look the same in five years, let alone ten.

1st prediction: Generated images will be photo-realistic

We can already see that some AI generated images are already tricking people into thinking they are real. I predict this will be even harder to distinguish in 2030, if at all.

More specifically, we can expect future models to excel at:

• Detail: faces, hands, textures, reflections, etc
• Multi-image consistency
• Complex scenes

The quality of the output will only increase in the future as the training of AI models and the power of the computers improve. Yes, that does create some uncomfortable questions around authenticity. But let’s deal with that later on.

Prediction #2: AI art will no longer be just images

Midjourney at present is an image-based service. I suspect that the future will be multimodal: images, videos, animations and possibly even interactive work. Think being able to generate a short movie.

Including movement, sound and editing. I’m seeing AI tools already moving in this direction, as they integrate with other media types.

It feels a little futuristic, but AI-generated art felt like that a few years ago too.

Prediction #3: Creativity will become more about ideas than execution

This is already underway, but it will probably accelerate.

If anyone can create top-notch imagery in a split second, the emphasis shifts from execution to idea generation. Ideas, narrative, innovation.

According to surveys, AI adoption is forcing professionals to spend more time on strategy and creativity and less on production, HubSpot reports.

Honestly, that’s both exciting and a little daunting. Because thinking of good ideas? That’s not as easy as it sounds.

Prediction #4: New careers will emerge (and old ones will adapt)

Whenever technology advances, so do occupations. AI art is no exception. We’re already seeing positions like:

• Prompt engineer
• AI-assisted designers
• Creative directors for AI workflows

By 2030, these positions may become more established…and widespread.

The states that AI will both replace and generate jobs, especially in the creative and knowledge-based sectors. So, it is not only about loss, it is about transformation. Messy, but not necessarily negative.

Prediction #5: Ethical concerns will be raised

This one is inevitable. The better AI-generated art gets, the louder the discussions around copyright, ownership and authenticity will become. Who owns the rights to an AI generated image? What about copying styles? How do we tell what is real and what is not?

These are not hypothetical questions anymore. They are being discussed, and they will probably influence the way Midjourney and similar tools will develop. According to experts, regulation and ethics will become increasingly important as the use of AI expands .

And, to be frank, this is where it gets ugly. Technology is fast. Regulation… less so.

So, what will 2030 be like?

I would describe it as: more powerful, more convenient, more nuanced. Midjourney, and similar instruments, will probably be:

• Faster
• More accurate
• More woven into our routines

But also more controversial, more critical, more… human, in that it will bear the imprint of our priorities and decisions. Perhaps that is the lesson. The future of AI art is less about the quality of the output and more about how we decide to engage with it.

Ultimately, the instrument will only be able to take us so far. What we do with it is still up to us.

How many images are being created every day?

Millions. This is the number of images (from concept art, to ads) being created every day.

How long does it take to create an image?

Less than a minute. This is the typical time it takes to create a finished image. In contrast, a human artist might take anywhere from a few hours to a few days to create the same image.

How long are people’s prompts getting?

Really long. People are writing enormous prompts to customize their results. For advanced creators this might include references, lighting information, composition, etc.

How many attempts does the typical person take?

Most users generate multiple variations before settling on a final image. This is consistent with how other design tools are used.

Are people using the upscale (enhancement) features?

Yes. The majority of our users rely on at least one upscalse. Many rely on multiple.

Commercial vs. personal usage?

A significant portion of generated images are used for commercial purposes, including marketing, branding, and content creation. Others use the platform for personal projects or experimentation.

Is this cheaper than a designer?

Access to Midjourney is priced at a fraction of the day rate of a professional designer. But, for more complex applications, humans will be needed. Cost savings are a major factor for adoption.

Number of paying customers of Midjourney

Midjourney’s revenue model is based on paid subscriptions. Many people upgrade to Midjourney’s paid plans to generate more images and get them faster. Number of paid users will give us an idea about the demand for the tool. Monetization will grow as the user base grows.

Discord based interface

Midjourney is built on top of discord, and that is how you will use it. The use of discord will give you a different experience. You will use the tool in discord public channels. That way, you will see what others are generating. It gives a lot of room for experimentation and learning. You will learn from others and that will speed up your learning curve.

Learning curve

Midjourney has a good learning curve. The community is playing a big role in learning. People are sharing their prompts, techniques and generated images, etc. So, the learning for beginners will be much easier.

Copyright issues

Copyright laws are yet to be established for AI generated images. Users typically have rights to their creations, but legal frameworks are still evolving. There is a lot of discussion on the topic.

Ethical concerns

AI generated images can be used for unethical purposes. For example, deepfakes. Deepfakes are a form of manipulated media wherein a person in an existing image or video is replaced with someone else’s likeness. Though deepfakes are AI generated, the use of AI is not the issue here. Platforms are working to avoid misuse.

Frequency of correct output

Midjourney may not always give you the expected output. Sometimes the hands will be distorted. Sometimes, the image will be unrealistic. There are ways to avoid such issues. With better prompts, and advanced models, we can avoid them.

Improvement in rendering quality

You will see an improvement in rendering quality over time. With every new version of the model, you will see an improvement in quality. The output will have better details, lighting, and composition. Improving quality is one of the reasons people are coming back.

Time saving for artists

An artist’s job is faster with Midjourney. With traditional methods, it may take a few days to generate an image. With Midjourney, it will take a few minutes. This is one of the reasons for its adoption.

AI generated images vs stock photos

For some use cases, AI generated images are replacing stock photos. AI generated images give more control and originality over stock photos. This may impact stock photo industry. But, both will co-exist.

Integration with other tools

Most of the users are integrating Midjourney with other tools. For example, with adobe photoshop for post processing. With this, the final output will be improved. AI will take care of the ideation and traditional tools will take care of post-processing.

Prompt engineering

Writing a prompt is a skill now. Based on how you write the prompt, the output will vary. Skilled prompt writers can generate better images. That is why, a new role called prompt engineer has evolved. This is a new skill in town.

Too many AI generated images

Too many AI generated images are being generated. Standing out of the crowd will be a challenge. With a number of AI generated images, originality and uniqueness will be a challenge. It is like any other content in the digital world.

Impact on creative jobs

AI image generation will impact a few jobs in the creative industry. Some jobs will be lost and some new jobs will evolve. Humans and AI will work together in the creative industry.

Conclusion

All the numbers, trends, and patterns considered, what can we conclude? Midjourney is not a tool. It’s a movement. And like all movements, it’s hard to grasp while you’re in the thick of it.

We know it’s growing, we know it’s becoming more international, we know it’s having a widespread impact on various industries, but that doesn’t quite explain what it’s like to actually use Midjourney, with all its experimentation and little victories and sometimes throwing your hands up in the air because the results are almost, almost, almost right, but just… not… quite.

But perhaps that’s the nature of this particular beast. We don’t know how it will all play out yet, because it hasn’t played out yet.

Art is becoming devalued. Creativity is expanding. What is possible by a human vs. a machine is blurring and blurring and blurring. To some, this is a huge plus. To others, it’s a huge negative. And to most of us, it’s a mixed bag, as we muddle through. So where does this leave us?

Well, in a bit of a murky spot, I suppose, but a potentially exciting one, too, depending on how you choose to approach it. Because even if we don’t know what this will all ultimately mean, one thing seems pretty clear: creativity is not going away. It’s shifting. And whether you like that or not, it’s something we can no longer ignore.

Sources:

• DemandSage
• Business of Apps
• EarthWeb
• Coolest Gadgets reports
• Statista
• HubSpot
• Stable Diffusion
• McKinsey’s analysis of generative AI’s economic impact
• IBM
• World Economic Forum