Midjourney's Co-Star deal is less about astrology than control. A company born inside Discord is now buying the kind of mobile habit it never had to build.
Midjourney acquired Co-Star, the astrology app with more than 30 million registered accounts, according to a Bloomberg report published July 23. Terms weren't disclosed. The move is plain. Midjourney is trying to own more of the consumer relationship instead of borrowing it from Discord.
That is a sharp turn for a company that became one of AI's strangest businesses by refusing the usual startup script. It launched publicly as a Discord bot in 2022, charged users directly, stayed self-funded, and avoided the venture-capital race that shaped most of generative AI. But the company no longer looks quite as tiny as the old mythology suggests. Midjourney's own site now describes it as a community-funded lab of 60 people, and Forbes recently listed the same headcount while saying PitchBook put its 2024 revenue at $300 million.
David Holz has been clear about why Discord worked. In a 2022 Stratechery interview, he said Midjourney's team first used Discord internally, then watched users bounce prompts and ideas off one another in real time. The social setting made the tool feel alive. He went all in. Co-Star tests a different question: what happens when Midjourney's next social environment isn't a server full of prompt nerds, but a phone app built around daily personal ritual?
Distribution is the real product #
Co-Star isn't a creative tool. It doesn't generate images or help you design a flyer. What it has is a mobile-native audience. Co-Star's App Store listing shows more than 205,000 ratings, daily horoscopes, friend compatibility features, and paid add-ons such as Pro-Star monthly at $8.99. Insider previously reported that Co-Star had 30 million registered accounts and an average user age of 18 to 25. That is not a feature list. That is habit.
The app also isn't some vague wellness brand with a zodiac skin. Its own listing says it uses NASA data, real human astrologers, and AI technology to produce birth charts and daily horoscopes. You don't have to believe in astrology to see the consumer product lesson. Millions of people have already been trained to open Co-Star because it tells them something personal and a little strange - something worth sending to a friend.
Midjourney needs that kind of front door because AI image generation has become crowded fast. The 26.8% share cited in several market estimates appears to be a user-preference or image-generator market estimate, not a clean audited global revenue figure, so you shouldn't treat it as gospel. Still, the direction is obvious enough. OpenAI's image tools, Adobe Firefly, Stable Diffusion forks, Flux-based tools, and consumer design suites have made basic image generation cheap and familiar. When the image model stops feeling rare, distribution starts doing more work.
Canva understood this before most AI companies did. Its own press materials now put it at more than 265 million monthly active users and $4 billion in annualized revenue, while Inc. reported it ended 2025 with 31.2 million paid users. Canva wins because non-designers already know where to go when they need a poster, pitch deck, Instagram tile, you name it. Adobe responded with Express because it couldn't leave that market alone. The pattern is consistent: the company that owns the daily workflow gets more chances to sell the next tool.
This will not be automatic #
Midjourney's Discord base was brilliant in 2022. It was useful. Discord's own case study says around seven million servers have installed the Midjourney app, which tells you how far the bot traveled beyond the official community. But Discord still skews toward enthusiasts, creators, and technically patient users. Co-Star's audience is broader, younger, and far less likely to think of image prompting as a hobby.
That gap is the risk. There is no obvious reason someone checking a birth chart every morning wants to make AI images in the same app, or even in the same product family. The integration has to be earned. If Midjourney simply bolts a visual generator onto astrology, users will feel the join immediately.
What makes the deal plausible is Midjourney's financial position. The company is self-funded and profitable by multiple public accounts - and still far leaner than the AI labs spending billions on staff, chips, and distribution deals. A company with that structure can make an odd consumer acquisition and let it breathe. A venture-backed startup often can't. The clock is different when nobody is marking up your next round.
The harder question is whether Midjourney's identity can stretch from image generation into a broader consumer app suite. Co-Star built its following on a specific emotional register: personal, predictive, a little eerie, and written for the individual user. Midjourney's best images work in a related way. You type something private or strange, and the machine returns an object that feels like it came from your own head.
Frankly, that is the connection Midjourney is buying. Not astrology. Not horoscopes. A daily consumer relationship that feels personal enough to survive outside someone else's platform. Discord was a brilliant shortcut, but it was never going to be a permanent home.
Also read: CATL's energy storage business grew 88% in six months as AI data centers become its next major market • Congress introduces an AI kill switch bill after an OpenAI model hacked its way out of a sandbox • Unitree's AS2-W wheel-legged robot scales cliffs on reinforcement learning and undercuts Spot by a mile