Midjourney’s Co-Star Acquisition Signals a Bigger Consumer AI Bet

Midjourney made its first acquisition despite lacking a standalone app, buying social astrology service Co-Star during a much broader expansion beyond image generation. The unusual pairing drove TechCrunch Verge coverage because it combines an AI image lab, personalized horoscopes, and an established mobile audience.

That framing makes the deal sound like an eccentric side project. It is more useful to read it as an organizational purchase. Midjourney is acquiring product leadership, consumer-app experience, and a daily habit shared by millions of users.

Co-Star founder Banu Guler and roughly two dozen employees are joining Midjourney. Co-Star will remain under Guler’s control, while she helps build a broader product and design organization around Midjourney’s expanding portfolio.

That structure creates the central tension. Midjourney became a major consumer AI brand through a Discord-centered experience, while Co-Star grew through a dedicated mobile product built around identity, relationships, and recurring engagement.

The acquisition therefore tests two different routes into consumer AI. One begins with generative capability and searches for products. The other begins with a personal ritual, then uses algorithms to deliver content at scale.

What Midjourney Actually Bought

Midjourney bought an operating consumer product and its product team, not merely an astrology content catalog.

Midjourney announced the acquisition on July 23, calling Co-Star its first purchase. The deal reportedly closed during spring 2026, although neither company disclosed its financial terms.

Co-Star offers personalized horoscopes, birth charts, compatibility assessments, and advice. Users enter their birth time, date, and location, then add friends and compare astrological profiles.

The app combines human-written interpretations with software that matches those interpretations to planetary positions. It also uses natural-language AI for more conversational features, including personalized answers to open-ended questions.

According to acquisition details, Co-Star has about 4.3 million monthly active users. That estimate comes from Sensor Tower rather than a new figure released by either company.

Midjourney’s announcement emphasized Guler’s role more than Co-Star’s traffic. Founder David Holz said he had known Guler for more than five years and regularly discussed reflection and decision-making with her.

Holz said Guler and her team would help build a broader organization around product and design. He also said Co-Star would remain entirely under her direction, with expanded resources from Midjourney.

That language matters because it separates the acquisition from a conventional product consolidation. Midjourney did not announce that it would absorb Co-Star, rebrand it, or replace its existing system with a Midjourney model.

Instead, Midjourney appears to be adding a semi-independent consumer operation. Co-Star retains its identity while its team gains responsibilities across the buyer’s wider collection of projects.

The timing strengthens that interpretation. Holz said Midjourney plans to reveal half a dozen ambitious projects during the next six months. He presented Co-Star as an early part of that expansion rather than an isolated experiment.

Midjourney has already moved outside its familiar image generator. The company has discussed medical research, including imaging applications, and has signaled interest in physical spaces and other unconventional projects.

Bloomberg also reported that Midjourney is building its first standalone image-generation app. That project would require skills Co-Star’s team already practices, including onboarding, notifications, mobile interaction, retention, and consumer product design.

Midjourney’s traditional interface grew around Discord, where users submit prompts and watch generated images appear inside shared channels. The arrangement helped build community, but it also made Midjourney dependent on another company’s platform.

A dedicated app would give Midjourney greater control over discovery, identity, moderation, payments, and user relationships. Those are product challenges, not simply model-training problems.

The official announcement supports that reading. It describes Guler’s team as help for Midjourney’s entire product and design organization, not as staff assigned only to horoscopes.

The event triggering the TechCrunch Verge discussion is therefore bigger than its strangest detail. Astrology attracts attention, but experienced consumer-product leadership explains the acquisition more clearly.

Midjourney gained a live mobile product, a recognizable voice, and millions of recurring users. It also gained a team that understands how an abstract algorithm becomes a personal daily experience.

Why the TechCrunch Verge Story Is Really About Product Design

The deal exposes Midjourney’s central weakness: impressive generative models do not automatically produce durable consumer products.

Midjourney established its brand through outputs. Users supplied prompts, generated images, shared results, and learned techniques from each other. The model’s visual quality did much of the selling.

That loop works well when novelty and creation drive usage. It becomes less dependable when a company wants daily engagement, broader audiences, or products unrelated to image generation.

Co-Star developed the opposite skill set. Its core attraction is not an obviously superior computational model. It is a designed relationship among personal data, sharp writing, notifications, friendship features, and repeated self-reflection.

The app asks for birth details, generates a natal chart, and lets users compare compatibility with friends. A natal chart is astrology’s map of planetary positions at a person’s birth.

Co-Star’s system then chooses relevant interpretations for each user. Its product explanation says human contributors work with planetary data and natural-language AI to deliver personalized context.

The company also says it adjusts surfaced content according to behavior inside the app. If someone repeatedly examines particular chart areas, Co-Star can prioritize related material later.

That personalization creates a feeling of continuity. The product appears to remember what concerns the user, even though its recommendations originate from a structured content and rules system.

This distinction is important for consumer AI. Generating an answer is only one part of a product experience. The harder work often involves deciding when to speak, what to remember, and how to earn another visit.

Co-Star has spent years building that experience around questions of identity and relationships. Users do not merely request an output and leave. They compare profiles, interpret notifications, and discuss results with friends.

In 2021, Co-Star said it had surpassed 20 million downloads without conventional marketing. It also said one quarter of American women aged 18 to 25 had downloaded the app.

Those numbers were company claims from an earlier growth period, not current active-user measurements. Still, they show why Midjourney would value Co-Star’s understanding of organic distribution and social sharing.

An image model can produce something visually striking. A social consumer app must turn output into conversation, identity, and routine. Co-Star’s popularity came from packaging its content as something people could exchange.

The app’s tone was central to that packaging. Its compact, sometimes severe notifications were easy to screenshot and share, giving its algorithmically assembled content a recognizable editorial identity.

Midjourney faces a similar challenge as generative media becomes common. Visual quality still matters, but model improvements alone become harder for ordinary users to distinguish.

A standalone app needs more than a prompt box. It needs clear workflows, understandable controls, reliable account management, reasons to return, and an identity separate from its underlying model.

Co-Star’s team has experience across those layers. That background explains why Guler will work across Midjourney’s portfolio while continuing to direct her existing app.

It also places pressure on other model companies pursuing consumer products. OpenAI, Google, Adobe, and independent creative platforms increasingly compete through distribution and workflow, not raw generation alone.

Midjourney cannot rely indefinitely on users arriving through Discord and adapting themselves to its interface. A broader product portfolio requires Midjourney to meet users inside purpose-built experiences.

The TechCrunch Verge keyword may reflect readers searching across both outlets for confirmation. Yet both reports point toward the same strategic signal: Midjourney is becoming a multi-product company.

That transition changes how the business should be judged. Future success depends on whether Midjourney can create coherent consumer experiences, not how many unrelated research projects it announces.

Co-Star’s Daily Habit Meets Midjourney’s Generative Engine

The primary contest is capability-led AI versus habit-led consumer software, and Midjourney is trying to combine them.

Capability-led companies begin with a model that can perform impressive tasks. They then search for interfaces, audiences, and recurring situations where those abilities become valuable.

Habit-led products begin with a repeated human need. They use technology to make that behavior easier, more personal, or more socially rewarding.

Midjourney clearly followed the first route. Its image generator established technical credibility, attracted creative communities, and produced a globally recognized visual style.

Co-Star followed the second route. It took an old interpretive practice and delivered it through a smartphone ritual involving identity, relationships, and daily uncertainty.

Astrology is not scientifically validated as a method for predicting events or personality. That limitation does not prevent astrology products from functioning as tools for entertainment, conversation, or personal reflection.

Co-Star founder Banu Guler has argued that the practical question concerns astrology’s effects as a reflective framework. In a 2021 founder interview, she compared the experience to talking about yourself with close friends.

That framing reveals what Midjourney may find useful. Co-Star does not sell access to astronomical computation. It sells an intimate interface for considering emotions, choices, and relationships.

Its “Ask the stars” feature makes that structure more obvious. Users can submit questions concerning love, work, relocation, self-sabotage, or social connection.

Co-Star says answers draw from a database developed by poets, astrologers, and technologists. Its system parses a question, identifies relevant chart placements, then assembles an interpretation.

This is generative software inside a bounded product. The response has a defined purpose, an established voice, personal context, and limits around questions the service considers unsafe.

A general chatbot can answer similar questions without astrology. However, generality does not guarantee an experience users recognize, trust, or discuss with friends.

Co-Star supplies a frame. Every answer passes through a consistent language, visual system, and theory of personalization. That consistency can make a product feel more coherent than an open-ended assistant.

For Midjourney, the lesson extends beyond horoscopes. Consumer AI becomes more durable when the model supports a recognizable activity instead of demanding that users invent one.

A dedicated Midjourney app could apply that principle to creativity. It might organize generation around projects, social interaction, personal taste, or recurring creative goals.

That outcome remains speculative because Midjourney has not detailed the app. The acquisition provides organizational evidence, not a confirmed feature roadmap.

Still, Guler’s cross-company role suggests Midjourney wants more than engineering support. It wants judgment about how products speak, guide behavior, and become part of everyday life.

The deal also raises a broader question about personalization. Midjourney’s models can learn visual preferences, while Co-Star interprets personal data and past behavior through an editorial system.

Combining those capabilities could produce more adaptive creative tools. It could also create experiences that feel unusually personal without offering users enough clarity about their operation.

Many knowledge tools face a related design problem. A useful personal knowledge system must provide context without pretending that stored patterns fully define a person.

Co-Star’s approach shows both the attraction and danger of that promise. Users enjoy content that seems specific to them, but specificity can encourage misplaced confidence in automated interpretation.

Midjourney must learn the right lesson. The value lies in Co-Star’s product discipline and recurring engagement, not in treating personalized output as inherently accurate.

This is where the acquisition becomes a genuine strategic test. Midjourney must combine model capability with product habit while preserving boundaries between reflection, entertainment, and dependable advice.

Personalization Creates a Trust Problem

Co-Star gives Midjourney a proven personalization engine, but it also brings sensitive data, subjective claims, and heightened responsibility.

Co-Star asks users for their exact birth time, date, and location. Users may also connect phone numbers, add friends, and submit private questions about intimate decisions.

The company says it does not access texts, microphones, or third-party tracking data. It also says it does not sell user data and uses in-app behavior to select which content appears.

An acquisition naturally makes users ask whether those practices will change. Midjourney’s announcement did not describe data-sharing rules, integration plans, or new privacy controls.

It also did not explain whether Co-Star accounts will remain technically separate from Midjourney services. Continued operational independence does not automatically answer questions about infrastructure or data governance.

These uncertainties deserve attention because personalization can blur the line between engagement and influence. A generic horoscope is easy to dismiss, while a tailored answer can appear more authoritative.

Co-Star explicitly warns users not to follow the stars blindly. It says astrology describes conditions but cannot override individual choice.

That disclaimer is useful, yet the product also invites questions about leaving partners, changing careers, and moving countries. Those situations can carry emotional and financial consequences.

The risk is not that Midjourney suddenly endorses astrology as science. The risk is that increased resources make persuasive automated advice more available before accountability mechanisms become equally visible.

Co-Star says it refuses questions judged unsafe, harmful, or toxic. It also directs users in crisis toward mental health resources rather than trying to answer every prompt.

Those safeguards resemble controls used by conversational AI services. Their effectiveness depends on classification quality, policy consistency, and careful handling of ambiguous language.

Midjourney has not explained whether it will contribute models to Co-Star’s question-answering system. It has also not announced new safety testing or independent evaluation.

That silence does not prove an unsafe integration. It simply means reports should not assume that the acquisition immediately improves Co-Star’s AI or its protections.

There is also a product-quality question. Co-Star’s personalization relies partly on human-created interpretations attached to calculated astrological conditions.

Users sometimes treat the resulting text as uniquely generated for them. The underlying process can instead involve selecting and combining reusable pieces according to a structured system.

That difference is not inherently deceptive. Many recommendation products work by matching users to existing content. Problems emerge when the interface encourages certainty beyond what the system supports.

Criticism from astrology communities adds another layer. Some practitioners object to Co-Star’s interpretive choices, tone, and simplification of astrological traditions.

Other users value precisely those traits. They prefer an accessible, social experience over the complexity of a consultation with a professional astrologer.

Competitors occupy different positions. The Pattern focuses on personality and relationship patterns, while Sanctuary has emphasized access to human astrologers and tarot readers.

CHANI presents another human-led approach, pairing astrology content with a distinct editorial voice. These products show that AI is not the only route to personalization within the category.

Co-Star’s position combines automation, human writing, social comparison, and a sharp brand voice. Midjourney could strengthen that combination, but additional generative capacity will not resolve disagreements about accuracy.

The acquisition may also affect user trust in Co-Star’s independence. Guler retains control, yet Co-Star now belongs to a company pursuing several experimental projects.

Users will reasonably watch for changes in permissions, recommendations, account connections, and content generation. Clear explanations would matter more than broad assurances about creativity or beauty.

The same standard applies to Midjourney’s future standalone app. If personalization becomes central, users need understandable controls over what the system retains and how it shapes recommendations.

The TechCrunch Verge framing treats astrology as the surprise. The more consequential issue is whether Midjourney can govern highly personal consumer experiences with the same ambition it brings to generative media.

Midjourney Now Has to Prove This Is a Strategy

Three signals will show whether the Co-Star purchase created a coherent consumer AI strategy or merely expanded Midjourney’s collection of experiments.

The first signal is Midjourney’s standalone app. Bloomberg reported that the company is building one, but its launch date, structure, and relationship with Discord remain unclear.

A polished app with distinctive workflows would strengthen the organizational thesis behind the acquisition. It would show that Co-Star’s product experience is helping Midjourney control its consumer relationship.

A basic wrapper around existing generation commands would weaken that thesis. Midjourney does not need another place to enter prompts; it needs a product designed around sustained creative work.

The second signal is Co-Star’s product direction. Users should watch for new generative features, revised recommendations, broader media creation, or visible changes to the app’s editorial process.

A measured release with clear explanations would support Midjourney’s promise of expanded resources under Guler’s control. Sudden integration without transparent boundaries would raise concerns about independence and trust.

Changes to privacy policies will be especially important. They can reveal whether account information, behavioral data, or personalized content begins moving across company products.

The third signal is Midjourney’s next group of projects. Holz promised half a dozen ambitious announcements within six months, creating a concrete window for judging the company’s direction.

Projects connected by common product capabilities would suggest a deliberate portfolio. They might share design systems, personalization methods, creative infrastructure, or a consistent theory of user experience.

A collection of unrelated announcements would suggest something looser. Research breadth can produce valuable discoveries, but consumer businesses still need recognizable reasons to exist together.

Co-Star therefore becomes an early test of Midjourney’s management model. The company must support an independent product while asking its leaders to improve a broader organization.

That arrangement can work when responsibilities are clear. It becomes difficult when a small team must maintain its own service while advising several new initiatives.

The number of employees involved makes execution worth watching. Roughly two dozen Co-Star staff members are joining Midjourney, according to published reports.

That is meaningful product talent, but it is not an unlimited organization. Midjourney will have to prioritize where Guler’s team can make the greatest difference.

It must also retain the qualities that made Co-Star attractive. A larger parent company can provide resources, yet additional layers can weaken the voice and speed of a focused consumer app.

For competitors, the immediate pressure is limited. The Pattern, CHANI, Sanctuary, and other astrology products do not suddenly face Midjourney-generated horoscopes with demonstrated superiority.

Their longer-term challenge concerns distribution and product investment. A well-funded AI lab can give Co-Star access to engineering, media generation, and design resources that independent competitors may struggle to match.

Creative AI competitors face a different pressure. Adobe, OpenAI, Google, and specialized design platforms already understand that models must live inside usable workflows.

Midjourney’s acquisition signals recognition of the same reality. Its advantage in visual generation will matter less if other companies offer clearer, safer, and more integrated products.

Readers searching the TechCrunch Verge story should therefore look past the novelty of astrology. This is Midjourney’s first formal attempt to acquire the consumer-product capabilities it did not build through Discord.

The purchase does not prove that Midjourney can operate a portfolio, launch a successful app, or responsibly expand personalized advice. It does show that company leadership recognizes those challenges.

Over the next three months, watch the standalone app, Co-Star’s policies, and the shape of Midjourney’s promised projects. Together, those signals will reveal whether this acquisition connects capability with a durable consumer habit.

The useful question is not whether an image generator belongs with an astrology app. It is whether Midjourney can turn an unusual pairing into products people understand, trust, and choose repeatedly.

Follow those product changes instead of treating the acquisition as a one-day curiosity. The next release will tell us more than the stars can.