Bootstrapped Midjourney Buys Astrology App Co-Star

Midjourney, the AI image lab that has famously refused venture capital, has made an acquisition of its own. The company has bought Co-Star, the venture-backed social astrology app, along with its roughly two-dozen-person team, in a deal first reported by Bloomberg. Neither side disclosed the price.

It is an odd pairing on its face. Midjourney is known for the image and video generators that run inside its Discord community and web app, while Co-Star built a following by turning NASA orbital data and human-written horoscopes into blunt daily push notifications, sold through a paid premium tier. What connects the two is less the product than the people. TechCrunch reported that Co-Star’s team of about two dozen has moved to Midjourney, and that the app claims roughly 4.3 million monthly active users.

An unusual acquirer

Most AI deals this cycle have been paid for with investor money, which is what makes the buyer here stand out. Midjourney has rejected venture capital since David Holz founded it in 2021, and Holz has said its revenue moved well past $200 million a year while the company stayed profitable. That profile is rare in generative AI, where most competitors raise and burn outside capital to fund model training and compute.

It also changes how to read the transaction. A self-funded, profitable company almost certainly paid for Co-Star out of cash flow rather than a fresh round, and it answered to no board in deciding to spend that money on an astrology app. Because Midjourney has never raised, no outside valuation has ever been set on it, and no investor had a vote on the diversification.

For Co-Star, the deal is an exit. The app raised a $5.2 million seed round in 2019 from investors including Maveron and the Female Founders Fund, then a Series A led by Spark Capital in 2021. Founded in 2017 by Banu Guler, it became one of the most recognizable consumer astrology brands, passing 20 million downloads by 2021. Its backers now collect a return of undisclosed size, and Guler’s team lands inside one of the most profitable independent companies in AI.

A talent deal dressed as a product buy

The more telling detail is what Midjourney lacks. Despite its scale, it has never shipped a standalone app; users reach it through Discord or a browser dashboard. Co-Star, by contrast, spent years building, shipping, and charging for a polished mobile product. Bloomberg reported that Midjourney is now building its own apps, and the purchase reads at least partly as a way to buy the consumer-app talent the lab has never had in-house.

Because the terms are undisclosed and Co-Star is a real product with millions of users, this is not a pure acqui-hire. But the shape is recognizable: a small, venture-funded consumer team joining a much larger, cash-rich buyer that wants what they already know how to build. That puts Midjourney in familiar company. OpenAI has used similar talent-focused acquisitions to absorb specific teams, and Meta’s $2 billion Manus purchase was widely read as buying a capability it could not build quickly enough itself. Midjourney’s version is smaller and stranger, but the logic matches: acquire a team that has already solved a problem you have not.

Casting a wider net

The astrology app also fits a broader pattern of Midjourney reaching well beyond image generation. The company recently launched a medical-imaging division and has floated plans for a spa, and Holz has described an ambitious slate of new consumer products. A working app with a loyal base and a subscription model gives Midjourney another distribution surface to test those ambitions against.

The pairing is also a small marker of where consumer AI is heading. Co-Star was an early example of a mainstream app blending machine-generated text with human editing, years before chatbots made that combination ordinary. For Midjourney, buying an established brand rather than growing one signals the same impatience showing up across the sector, where the fastest route to users and product talent is increasingly to acquire both at once.

What Midjourney has not offered is a plan. It has not said whether Co-Star will keep running on its own, fold into a future Midjourney app, or simply serve as an on-ramp for the team it just hired. For a company that built its identity on independence and patience, waiting years while rivals raised billions, the sharper signal is that it is now willing to buy growth it once insisted on building itself.