The bootstrapped AI image lab is building MRI-grade scanners hidden inside luxury spas, with no investors and no full FDA sign-off yet.
Midjourney, the profitable AI image company with no outside investors, is entering healthcare. Its new division, Midjourney Medical, is building a 60-second full-body ultrasonic scanner it claims rivals MRI at nearly 100 times the speed, with the first scanner-equipped spa set to open in San Francisco in 2027.
Picture this: you step into a shallow pool of golden light, sink slowly into warm water, and 60 seconds later you walk out with a 3D map of your insides accurate to a fraction of a millimeter. No clipboard, no waiting room. Just a soak that happens to scan you.
What Happened
The company best known for turning text prompts into striking AI art just announced something completely off-script: a move into medical imaging.
Midjourney Medical is building a machine the company describes as “as powerful as MRI, and as casual as a trip to the spa.” The pitch is simple and a little audacious. You want as much data about your body as quickly and cheaply as possible, and Midjourney wants to be the company that delivers it.
The framing they use is telling. They want to optimize for “megabytes per second per dollar” of information about your body. That is an engineer’s way of describing healthcare, and it tells you exactly how this team thinks.
The Backstory
Midjourney is not a typical tech company. Founded by David Holz, it took no venture capital, stayed profitable, and built one of the most popular AI image tools on the internet without the usual Silicon Valley funding playbook. The company calls itself a “community-backed research lab,” funded by everyday users rather than investors.
That independence matters here. A VC-backed startup answers to a board that wants a focused, fast path to returns. Midjourney answers to nobody but its users and its own founders, which is roughly the only structure under which a profitable AI art company decides to start manufacturing medical hardware.
The Plan
Here is how the machine actually works. You stand on a platform that lowers you into water at about 5 centimeters per second. As you descend, your body passes through a ring made of half a million tiny squares, each the size of a grain of sand. Every square doubles as a speaker and a microphone, firing ultrasonic waves through your body and listening for the echoes, like a choir of dolphins using echolocation.
The data volume is staggering. The system produces terabytes of data every second. Midjourney says that if you converted one second of scan data into HD video, you would need to watch 500 hours of footage. Thousands of computers then reconstruct that flood of sound into a detailed 3D map of your insides.
Then comes the genuinely clever part: the Midjourney Spa. Rather than selling people on going to a medical facility, Midjourney is wrapping the scanner inside a spa with hot tubs, saunas, cold plunges, and pools of golden light, open 24/7. The scan is, in their words, a “side-effect.” You go to relax, and you leave with a growing library of health data.
The roadmap is aggressive. The next 12 months go to refining hardware and algorithms. The first spa opens around the end of 2027. In 2028, the company moves to a custom-silicon third-generation scanner and starts scaling to more cities. The stated goal: a fleet of over 50,000 scanners worldwide by 2031, capable of a billion scans a month, enough to give monthly scans to a billion people.
The Business Model Angle
This is a textbook case of a company extending a hard-won capability into an entirely new market, and it is worth studying regardless of whether the scanner works.
Midjourney’s real moat was never “we make pretty pictures.” It was the underlying competence: massive compute, reconstructing images from messy data, and turning signal into pictures at scale. A body scanner that rebuilds a 3D image from terabytes of sound is the same core skill pointed at a new problem. That is how durable companies expand. They don’t chase adjacent revenue, they ask what their actual business model makes them uniquely good at, then find the biggest market that rewards it.
The spa twist is the sharper lesson. Selling preventive scans directly is hard, because healthy people don’t wake up wanting an MRI. So Midjourney changed the product category. By bundling the scan inside an experience people already want, they sidestep the adoption problem entirely. The scan becomes a byproduct of a Saturday you were going to enjoy anyway. Founders fighting low demand should note the move: sometimes the fix isn’t a better sales pitch, it’s a better wrapper.
And the funding structure is the quiet enabler. No investors means no pressure to justify a multi-year hardware bet against next quarter’s numbers. The unusual capital model is what makes the unusual product possible.
The Risk
Now the cold water, and there is plenty of it.
This is a far harder business than AI art. Software scales for nearly free. Building 50,000 physical scanners and operating real-world spas means manufacturing, real estate, staff, maintenance, and liability, all things a software company has never touched. The jump from generating images on a server to building bathwater MRI machines is enormous, and timelines like “billion scans a month by 2031” are the kind of projection that tends to slip.
Then there is regulation. Midjourney admits it is starting by offering body composition maps, not diagnoses, precisely because diagnostic capability normally requires FDA approval. The company plans to submit results to the FDA over time, but the most valuable promise, catching disease early, sits behind a regulatory wall it has not yet cleared.
And the headline claim deserves real scrutiny. Midjourney says that with enough early imaging, the world could avoid 30% of all deaths and 50% of all healthcare costs. Those are the company’s own projections, not validated outcomes, and they assume early detection reliably leads to better results. The medical reality is messier. Whole-body scans of healthy people can surface ambiguous findings that trigger anxiety, more tests, and unnecessary procedures, a known problem with broad screening. Bold vision, unproven math.
Quick Questions
What is Midjourney Medical actually building?
A full-body ultrasonic scanner that uses sound waves to reconstruct a 3D image of your insides in about 60 seconds, which the company claims is comparable to MRI but nearly 100 times faster.
Why is it being put in a spa?
Because people don’t line up for medical scans, but they do go to spas. Midjourney is wrapping the scanner in a relaxing experience so the scan happens as a side-effect of something you already enjoy.
Can it diagnose diseases?
Not yet. It is starting with body composition maps, not diagnoses, because diagnostic features need FDA approval. The company says it will submit results to the FDA to unlock more capabilities over time.
When can I actually use it?
The first Midjourney Spa is planned for San Francisco around the end of 2027, with a wider rollout to more cities starting in 2028.
The Bottom Line
Midjourney just reminded every founder that your moat is rarely the product you’re famous for. It’s the capability underneath it. The company turned “we reconstruct images from chaotic data” into a healthcare moonshot, and made it palatable by hiding the medicine inside a spa day. Whether the science holds up is a real and open question. But the strategic playbook, extend your core competence, reframe the category to kill the adoption problem, and use an unusual funding model to fund an unusual bet, is one worth stealing even if the scanner never ships.
