AI companion businesses (sometimes positioned as “virtual partners,” “AI Joi girlfriends/boyfriends,” or character chat) have moved from a novelty into a subscription-first consumer category with unusually strong engagement. The category sits at the intersection of three forces: (1) LLM quality and voice interfaces getting “good enough” for emotionally fluent dialogue, (2) consumer willingness to pay for personalization and private, always-on interaction, and (3) distribution via mobile stores where subscriptions and microtransactions are already normalized.
1. Market momentum: the “time spent” story matters more than downloads
The broader generative AI app economy is now operating at mass scale: 1.7B downloads in H1 2025 and roughly $1.87B in in-app purchase (IAP) revenue over the same period, with users spending 15.6B hours in GenAI apps. This matters for companion products because they thrive on habit formation. A utility chatbot can be “useful” once a week; a companion product wins when it becomes daily, emotionally sticky usage.
Europe is a meaningful growth engine in this broader wave: GenAI app downloads in Europe grew 51% half-over-half from H2 2024 to H1 2025. The implication for companion businesses is straightforward: even if only a small fraction of GenAI users convert into “relationship/character” use cases, the absolute addressable audience keeps expanding.
2. The companion segment: smaller revenue than assistants, but faster product iteration
Dedicated AI companion apps (as a defined subcategory) remain smaller than general assistants, but their growth profile is clear. Appfigures data (reported by TechCrunch) estimates:
- 337 active, revenue-generating AI companion apps worldwide; 128 launched in 2025 (as of August 2025).
- $82M in category revenue in H1 2025; on track for $120M+ for 2025.
- 220M lifetime downloads as of July 2025; 60M downloads in H1 2025 (up 88% YoY).
- $221M lifetime consumer spending as of July 2025; revenue per download rising from $0.52 (2024) to $1.18 (2025 YTD).
- Revenue concentration is extreme: top 10% of apps capture 89% of revenue.
That last point shapes strategy: “a companion app” is not automatically a business. The winners invest in retention mechanics, safety, and differentiated fantasy/personalization loops that justify ongoing subscription value.
3. How these companies make money: the standard playbook (and what’s emerging)
Most companion businesses monetize like premium consumer entertainment products rather than classic SaaS:
Core revenue streams
- Subscriptions (primary): paywalls around memory, longer messages, higher-quality models, voice calls, fewer ads, or “relationship modes.”
- Consumables / in-app currency: packs (often framed as “gems”) used for special interactions, gifts, images, or premium scenes.
- Tiered upsells: “Plus/Pro/Ultra” tiers, annual plans with discounting, and limited-time bundles.
- Creator economy (emerging): user-generated characters plus paid perks, marketplace-style revenue shares, or premium templates.
- Enterprise/licensing (select cases): white-label conversational characters for media brands, games, or communities (less common but strategically attractive).
Typical pricing in the market
- Replika’s App Store listing shows multiple price points, including monthly tiers at $7.99 and $14.99 plus annual plans and paid “gems,” illustrating the common “subscription + consumables” bundle.
- AI’s subscription page shows a yearly plan at $94.99 (presented as $7.92/month), highlighting the push toward annual prepay and predictable cash flow.
In practice, EU pricing often mirrors US pricing with local currency rounding and VAT handling; the commercial logic stays the same: annual plans reduce churn, while consumables monetize power users.
4. “Average profit in the EU”: a practical proxy (unit economics), plus a simple graph
Public data rarely discloses profit for this niche, so the most defensible way to discuss “average profit” is unit economics: estimated net profit per paying subscriber after platform fees and variable costs. A key constraint is app-store economics. For example, Google Play states that auto-renewing subscriptions are charged a 15% service fee from day one. Apple’s subscription commissions and EU-specific options can vary by program and structure, but subscription fees commonly sit in the 15–30% range depending on eligibility and tenure.
Below is an illustrative EU unit-economics graph: estimated monthly net profit per 10,000 paying subscribers at common monthly price points. Assumptions: 15% platform fee, and €4.50 per paying subscriber per month for variable AI + infra + support/moderation (costs vary widely by model choice and usage intensity).
ASCII “profit graph” (€/month, per 10k paying subscribers)
- €7.99 plan → ~€22,915 |
- €14.99 plan → ~€82,415 |
- €19.99 plan → ~€124,915 |
Interpretation: companion businesses can be very profitable at scale if they manage compute costs and churn. The same model can also be unprofitable at low price points if heavy users drive inference cost spikes, or if acquisition costs (paid UA) exceed LTV.
5. User portrait: who pays, and why they stay
Two patterns recur across the category:
- A) Younger, mobile-native cohorts anchor usage.
Replika’s App Store page (via a cited media quote inside the listing) notes its main users are aged 18–25, consistent with the category’s skew toward digitally social, always-on communication habits. - B) “Tech-savvy” and entertainment-adjacent personas over-index.
A Sensor Tower–based recap highlights higher overlap between GenAI assistant users and “tech-savvy” personas (e.g., crypto traders, students), while character/companion-style apps over-index with console gamers and “digital socialites.” This aligns with observed product behavior: roleplay and companion experiences compete with games and social media for leisure attention, not with productivity tools.
Motivations that drive conversion
- Low-friction companionship (always available, no social penalty)
- Roleplay/storytelling and identity exploration
- Stress relief and emotional regulation (often framed as “judgment-free”)
- Adult-oriented intimacy products (where allowed), usually monetized through stricter paywalls and higher ARPPU
6. The future of the industry: where moats will come from
Over the next 24 months, differentiation will likely concentrate in five areas:
- Multimodal realism: voice calls, better memory, images, and “presence” features that increase willingness to pay (and raise safety requirements).
- Retention engineering: habit loops, personalization, and content breadth; winners treat churn as the central KPI.
- Safety and compliance as a feature, not a cost: especially in Europe, where the AI Act has a staged implementation timeline (entered into force 1 August 2024; major provisions phase in through 2025–2027). Companion apps will need clearer age controls, transparency, and risk management—particularly where the product can influence user wellbeing.
- Cost advantage: model choice, caching, on-device inference (where feasible), and smart throttling for heavy users.
- Distribution strategy beyond app stores: community-led growth, creator ecosystems, and web funnels to reduce platform dependency (while remaining compliant with EU rules and platform policies).
