Midjourney has acquired Co-Star, a social astrology application with millions of regular users, in a transaction that gives the generative-AI company an established consumer-product team as it prepares to expand beyond its image- and video-generation services.

The acquisition is the first completed by Midjourney and was formally announced by the company on July 23. Financial terms were not disclosed. The transaction reportedly closed during the spring, meaning the public announcement followed several months of integration planning between the two privately held businesses.

Co-Star founder and chief executive Banu Guler will remain in control of the astrology application and continue directing its operations. She is also joining Midjourney as chief design officer, according to reports, with responsibility for helping establish a broader organization spanning design, product development and front-end engineering.

Co-Star’s team of approximately two dozen employees has joined Midjourney as part of the acquisition. The group’s experience operating a high-engagement mobile platform is likely to be central to the transaction’s strategic value. Midjourney has established one of the most recognizable brands in generative imagery, but it has not historically operated a conventional standalone mobile application comparable with the flagship consumer products offered by many other AI companies.

Midjourney founder David Holz said in the company’s announcement that Co-Star would remain under Guler’s direction while gaining access to substantially expanded resources and teams. He described the acquisition as part of the gradual development of a wider and more unusual research laboratory, rather than as an isolated purchase intended only to add an astrology business.

Holz also said Midjourney expects to announce roughly half a dozen ambitious projects over the following six months. The company’s public website already presents a product roadmap extending beyond image and video models, with planned software and hardware initiatives as well as a medical-imaging division. The Co-Star transaction places experienced consumer-app developers inside that expanding organization.

Reports indicated that Guler and her team will help Midjourney build its first dedicated applications, including an app focused on image generation. Such a product would give Midjourney a more direct relationship with consumers and reduce the importance of external platforms in the discovery, creation and sharing of AI-generated media.

Midjourney initially became widely adopted through Discord, where users entered text prompts and generated images inside public or private channels. The company has since developed a more complete web interface, including image creation, organization, editing, personalization and discovery features. A standalone application could make those capabilities more accessible to mainstream users unfamiliar with Discord or advanced prompting workflows.

That transition matters because competition in generative AI is increasingly moving from model performance to distribution and product experience. Improvements in image quality, prompt adherence and generation speed remain important, but several developers now offer systems capable of producing commercially usable media. Consumer adoption may therefore depend more heavily on interface design, personalization, community features and the ability to build habitual use.

Co-Star offers expertise in each of those areas. The application generates personalized horoscopes and astrological interpretations based on a user’s birth date, time and location. It also allows people to add friends, compare charts and receive recurring updates, turning an otherwise individual activity into a social and frequently revisited experience.

The company says its system combines astronomical positioning data, including information associated with NASA’s Jet Propulsion Laboratory, with a proprietary database of interpretations created by astrologers, writers and technologists. Natural-language software is then used to generate personalized content. The underlying astronomical data and the astrological conclusions drawn from it should be distinguished: astrology is not considered a scientific method for predicting events or personal outcomes.

Co-Star’s commercial model is also relevant to Midjourney. The application is free to download and earns revenue from optional purchases, including advanced readings, manually added friends and credits for asking personalized questions. According to Co-Star’s published frequently asked questions, all of its revenue comes from a la carte in-app transactions rather than advertising or the sale of user data.

That model gives Midjourney access to a team familiar with mobile payment systems, conversion funnels and monetization outside a conventional monthly subscription. Midjourney’s core generation service primarily relies on paid subscription tiers that provide varying levels of computing capacity, privacy and generation speed. Co-Star’s transaction-based approach could inform new pricing options for future consumer software.

The acquisition also provides Midjourney with exposure to an existing user base reportedly numbering about 4.3 million monthly active users. There has been no announcement that Co-Star accounts will be linked with Midjourney subscriptions or that the astrology app will directly incorporate Midjourney’s image or video models. Co-Star is therefore expected to continue operating as a distinct product unless the companies later disclose an integration plan.

Keeping Co-Star operationally separate may reduce disruption for its users while allowing Midjourney to retain the team responsible for its identity and growth. The arrangement resembles a talent-and-platform acquisition rather than a conventional product consolidation in which the acquired service is immediately absorbed, rebranded or discontinued.

Midjourney’s interest in Co-Star appears to extend beyond technical infrastructure. Holz said he had known Guler for more than five years and had regularly discussed how people reflect, make decisions and understand their past and future. Those themes are consistent with Midjourney’s stated ambition to develop technology around imagination, reflection, coordination, beauty and human flourishing.

The combination nevertheless brings together businesses with notably different public identities. Midjourney is associated with computational creativity, image synthesis and increasingly sophisticated generative models. Co-Star is a lifestyle and social application that presents astrological guidance in a distinctive, minimalist format. The immediate connection is therefore product design and personalized consumer engagement rather than a shared technical category.

For Midjourney, the acquisition could help close a gap between the quality of its models and the accessibility of its products. Generative-AI companies increasingly need multidisciplinary teams that combine machine-learning research with consumer design, mobile engineering, trust and safety, moderation, payments and customer support. Building those functions internally can take years, particularly for a research-focused organization accustomed to operating with a relatively small workforce.

Co-Star’s team has already managed many of those challenges. Its app handles sensitive personal inputs, recurring recommendations, social connections and questions involving relationships, careers and emotional distress. The company says it restricts responses to requests it considers unsafe or harmful and directs users to crisis resources when appropriate. Those safeguards could become relevant as Midjourney develops more personalized and interactive AI products.

The handling of user information will also attract scrutiny. Co-Star collects birth details and uses behavior inside the application to determine which astrological content to display. The company says it does not sell user data, does not access messages or microphones and does not track activity across unrelated apps or websites, although it uses analytics tools to assess product performance.

No changes to Co-Star’s privacy policies or data practices were announced with the acquisition. Any future connection between Co-Star profiles and Midjourney services would require clear disclosure, particularly because consumers may not expect information supplied to an astrology application to influence interactions with a generative-media platform. For now, there is no public evidence that such data sharing is planned.

The deal comes as Midjourney broadens its ambitions beyond its original creative tools. The company has established Midjourney Medical, a division developing a form of rapid whole-body ultrasound imaging that it calls Ultrasonic CT. It has also outlined prospective projects across software and hardware, suggesting a strategy that is considerably wider than the focused model-development approach followed by some AI laboratories.

This diversification creates both opportunity and execution risk. A self-funded company can pursue unconventional projects without the same quarterly pressures facing publicly traded technology groups, but managing consumer apps, advanced AI models, hardware development and medical systems requires very different regulatory, operational and commercial capabilities.

Co-Star may serve as a test of whether Midjourney can preserve the independence of specialized teams while building a coherent organization around them. Success would demonstrate that the company can move from a single dominant product toward a portfolio of consumer and research initiatives. Failure could leave management overseeing disconnected projects with limited technological or financial synergies.

The lack of disclosed financial terms makes it difficult to evaluate the acquisition from a valuation perspective. Co-Star previously raised venture capital, but neither company provided an acquisition price, payment structure or information about whether investors received cash, equity or a combination of the two. Midjourney is privately held and does not publish detailed financial statements.

The transaction is nevertheless notable because Midjourney has largely grown without following the aggressive fundraising and acquisition strategies seen elsewhere in the AI industry. Its first purchase indicates a willingness to use mergers and acquisitions selectively when an external team offers capabilities that would be difficult to reproduce quickly.

The most important measure of the deal will not be whether astrology becomes a significant new revenue category for Midjourney. It will be whether Guler and the Co-Star team can translate their expertise in mobile design, personalization and social engagement into products that make Midjourney’s generative technology easier to use and more deeply embedded in everyday consumer behavior.

Near-term indicators will include the release of a dedicated Midjourney app, changes in its mobile experience, new approaches to personalization and evidence that Co-Star can grow without losing its distinctive identity. Investors and competitors will also watch whether Midjourney makes additional acquisitions as it unveils the broader slate of projects promised by Holz.

For the wider AI market, the purchase reinforces a developing pattern: technical leadership alone is no longer sufficient to guarantee control of the consumer relationship. Model developers increasingly need recognizable applications, differentiated interfaces and communities that return frequently. By acquiring Co-Star, Midjourney has obtained all three—and a team now charged with applying those capabilities well beyond horoscopes.