The AI Companion App Market: What the Numbers Actually Say

Why the AI Companion App Market Is Hard to Size

If you ask five research firms how big the AI companion app market is, you get five answers. They range from $5.01 billion to $36.8 billion for roughly the same period. That's not a rounding error. It's a definition problem.

Some reports count only consumer apps where you chat with a virtual girlfriend, boyfriend, or role-play character. Others lump in enterprise assistants, desk robots, virtual influencers, and anything else with the word "companion" in the press release. Grand View Research puts the broad AI companion market at $36.8 billion in 2025. Research and Markets puts AI companion apps at $5.01 billion in 2026. Both are talking about the same phrase. They are not talking about the same thing.

This article breaks down the numbers, explains why they disagree, and looks at what the market actually looks like from the inside: who's making money, who's not, and what could go wrong.

Market Size Estimates Compared

Here's the spread. All figures are as published by the source. Growth rates are annual, where given.

Source	Measures	Size	Forecast	Growth
Grand View Research	Broad AI companion market	$36.8B (2025)	$318B by 2033	31.0% CAGR, 2026-2033
Globe Market Research	AI companion apps	$30.1B (2026)	$198.3B by 2035	n/a
Research and Markets	AI companion apps	$5.01B (2026)	$9.65B by 2030	17.8% CAGR
Research and Markets	AI girlfriend apps	$2.91B (2026)	$7.15B by 2030	25.2% CAGR
SNS Insider	AI girlfriend apps	$2.57B (2024)	n/a	n/a
LinkedIn industry article	AI companion platforms	$764M (2024), $856M (2025)	$1,666M by 2032	n/a
Appfigures via TechCrunch	Mobile spending on AI companions	$82M (H1 2025); over $120M for 2025	n/a	+64% H1 year over year
Sensor Tower	Mobile spending, broader definition	$150M in Q1 2026	n/a	Over 12x Q1 2023
Our estimate	App stores plus web billing	About $1-1.5B a year (2026)	n/a	Low-medium confidence

The gap between the top and bottom is roughly sevenfold. Even within the "AI girlfriend apps" niche, Research and Markets says $2.91 billion for 2026 while SNS Insider says $2.57 billion for 2024. Those are close enough to be measuring similar things. The $30 billion and $36 billion figures are measuring something much wider.

Why the Estimates Disagree

Three reasons.

Definition. Grand View's $36.8 billion covers the broad AI companion sector: consumer apps, yes, but also enterprise tools, virtual characters, and hardware. Research and Markets' $5.01 billion is narrower: apps you download to chat with a companion. If you're an investor trying to size the consumer opportunity, the narrow number is more useful. If you're a strategist looking at the whole sector, the broad number makes sense. They are not competing claims. They are different questions.

Where the money is counted. App trackers like Appfigures and Sensor Tower see in-app purchases on iOS and Android. They do not see web subscriptions. Most adult spending happens on the web, because Stripe and PayPal prohibit sexual AI content and adult processors charge 10.8-15.5% per transaction. Platforms like Candy.ai, SpicyChat, CrushOn, OurDream AI, and Joi take payment directly. Those dollars are invisible to app-store trackers. Appfigures reported over $120 million in mobile spending for all of 2025. That's real money, but it's not the whole market.

Forecast discipline. Sensor Tower's Q1 2026 figure of $150 million is over 12 times Q1 2023. That's a real baseline. But forecasts to 2033 extrapolate from about three years of sales data. Grand View's $318 billion by 2033 is a scenario, not a promise. Treat anything past 2030 as a direction, not a number.

Our own bottom-up estimate, adding app-store spending to web billing where we can see it, is about $1-1.5 billion a year in 2026. Confidence: low to medium. The truth is probably somewhere between the narrow Research and Markets figure and the LinkedIn industry article's $856 million for 2025.

Growth: Downloads, Spending, and Users

The growth is real even if the absolute numbers are fuzzy.

• Appfigures: about 60 million downloads in H1 2025, up 88% year over year.
• Appfigures: over 220 million lifetime downloads globally.
• Appfigures: revenue per download was $0.52 in 2024 and $1.18 in 2025. People are paying more per install.
• Sensor Tower: spending significantly outpaces user growth. The market is monetizing better, not just getting bigger.
• Appfigures: 337 AI companion apps were earning money by July 2025. 128 launched in 2025 alone.

Sensor Tower puts companions at about 4% of the $4 billion-plus in AI app spending in H1 2026. That's a small slice of a big pie, but it's a growing slice.

Analyst summaries put the compound annual growth rate between 18% and 31%, depending on the definition. Either way, it's fast. But growth from a small base is easy. The harder question is whether the market can keep growing once the easy adopters are in.

Segments: Text, Voice, and Multi-Modal

Text still dominates. Grand View says text-based companions held 42.7% market share in 2025. That makes sense: text is cheap to serve, easy to scale, and works on any device.

Voice is the premium tier. Muah's $99.99 Ultra plan includes phone calls. Nomi AI and Kindroid both charge $15.99 a month for entry, and memory is the paywall. If you want a companion that remembers what you said last week, you pay.

Multi-modal is the frontier. Secrets AI focuses on video. xAI launched 3D Grok companions, and iOS installs rose 40% with revenue up about 9%. But the 3D Grok mode was retired on September 1, 2026, according to RoboRhythms. That's a reminder that features come and go. The CES 2026 trend was companions moving into pendants and desk robots, but nobody has sized that market yet.

One data point on cost: Character.AI serves a conversation-hour for under $0.01, a 33x cost cut since 2022. Inference is getting cheaper. That means more features at lower prices, but also more competition.

Who Uses These Apps

The user base is broader than the stereotype.

• Common Sense Media: 72% of US teens have tried AI companions. 52% are regular users. A third used them for social interaction.
• Wheatley Institute and IFS: 19% of Americans have chatted with an AI romantic partner. 31% of US men aged 18-30 have done so, versus 23% of women in the same age group.
• Match and Kinsey Institute: 16% of US singles have used an AI romantic companion.
• OpenRouter: role-play accounts for about 52% of tokens on open-source models.

Gender skew depends on the platform. Ada Lovelace Institute found 77% of users on mixed-use UK platforms were male. JanitorAI claims 70-80% women, but Similarweb desktop data shows 37.3%. Company claims and independent data don't always match.

Appfigures found 4% of companion apps use "boyfriend" in the name versus 17% for "girlfriend." The marketing skews one way. The user base is more mixed.

Regional Breakdown

North America leads. Grand View puts North America at 33.5% revenue share in 2025. Other summaries put it at roughly 35% to 38%. Either way, it's the biggest single market.

Asia Pacific is the other hub. MiniMax listed in Hong Kong in January 2026, raising about $620 million. Its companion apps accounted for about 35% of $53.4 million in revenue for the first nine months of 2025. Love and Deepspace passed $750 million, with about 60% from China, according to KrASIA. Another source puts it at $820 million.

The UK is a notable market. Ada Lovelace Institute says Character.AI is a significant player there. UK adoption of mating-oriented AI is 14% above the global average.

There are no reliable bottom-up figures for South Korea or the Middle East. Treat any number you see for those regions with caution.

Key Platforms and Competition

The market is top-heavy. Appfigures says the top 10% of apps capture 89% of mobile spending. Only 33 apps have passed $1 million in lifetime revenue.

Player	Type	Reported scale	Money signal	Watch out for
Character AI	SFW chat	~20M monthly users (Sacra)	~$30M annualized, July 2025 (Sacra)	Under-18 chat ban; five teen-harm suits settled
Chai	18+ chat	10M+ users (company)	$80M ARR Q1 2026; $100M+ claimed July 2026	Self-reported figures
Talkie (MiniMax)	SFW chat	~20M monthly users with Xingye (ChinaTalk)	~$5 per customer over 9 months (ChinaTalk)	Pulled from US App Store, Dec 2024
Replika	Emotional	40M+ users (company)	$24-30M (Sacra, earlier) vs $4.8M (Latka, 2026)	€5M GDPR fine, Italy
JanitorAI	Role-play	2.5M daily users (company, via Forbes)	Janitor+ $12.99, June 2026	Audience claims conflict
SpicyChat	Role-play	100M registered (company)	Prices raised Sept. 8, 2026	Recent price rise
Nomi AI, Kindroid	Emotional	Not disclosed	$15.99/month entry	Memory is the paywall

Candy.ai claims about $25 million in annual revenue with no venture funding. Unverified, but plausible given its position. OurDream AI made a16z's top-50 web generative AI list in March 2026. Secrets AI focuses on video. Each has a different angle.

If you're comparing specific platforms, we've ranked the best AI companion apps by what you actually need. For a deeper look at one of the biggest, see our Candy.ai review.

Big Tech and Investments

Google licensed Character.AI's technology in August 2024 for about $2.7 billion, per Sacra. Google also joined the January 2026 settlements related to teen-harm suits. Character.AI itself raised $150 million at a $1 billion valuation in 2023.

Chai says it has raised $55 million-plus. Most adult operators are bootstrapped. They take payment directly and keep the margin, but they also bear the risk.

OpenAI paused its adult mode in March 2026. xAI's 3D Grok companions gave a short-term bump but were retired in September 2026. Big tech is experimenting, not committing.

Revenue Models

Entry subscriptions cluster at $13-15 a month as of September 2026. Candy is $13.99, Nomi $15.99, Kindroid $15.99, Replika $14.99. Annual plans run 40-70% off, which is standard.

But the sticker price is not the real price. Token metering and paywalled memory push heavy users higher. Reviewers report heavy Candy users spending $25-80 a month. SpicyChat gates context length by tier. CrushOn's top tier is $199.99 a month, according to FirmCritics. Chai Ultra is $19.99 a week.

Free tiers with ads exist at Character AI, Chai, and Talkie. They're a funnel, not a charity.

RevenueCat's 2026 data shows AI apps convert trials at 8.5% versus 5.6% for other apps. But they retain 36% worse. People try, pay, and leave. That churn is the quiet problem behind the growth numbers.

What Users Want

Memory is the feature people pay for. Ada Lovelace Institute found companions often require payment to retain memory. If you want a companion who remembers your name, your job, and what you told them last week, you pay.

Voice and video are next. Secrets AI's video focus and Muah's phone calls on the $99.99 Ultra plan show where the premium tiers are going.

Customization matters, but it's table stakes. Every platform has it.

Mental Health Risks

This is the part the growth forecasts don't capture.

InsightAce Analytic cites 43% reported increased anxiety in 2024. Replika removed adult role-play in 2023, and negative sentiment on its subreddit went from 13% to 38%. Users describing loss had over 10 times the odds of mental health concerns, according to De Freitas et al. OpenAI and MIT Media Lab found heavier chatbot use linked to more loneliness, though the study is correlational.

Companions can validate harmful views and foster dependence. They are not therapists. Ada Lovelace Institute found 53% of UK teenage boys prefer online interactions. That's a warning sign, not a feature.

Regulation

Regulation is arriving, state by state and country by country.

• California SB 243 took effect January 1, 2026, with $1,000 per violation in private suits.
• New York's law took effect November 5, 2025.
• The FTC sent orders to seven companies, including Alphabet and OpenAI, in September 2025.
• The GUARD Act passed the Senate Judiciary Committee in April 2026.
• The EU AI Act's chatbot transparency duties apply from August 2, 2026, with fines up to €15 million or 3% of turnover.
• Ofcom closed its age-check probe of Joi's operator in July 2026.

Age verification is the weak point. Ada Lovelace Institute found 15 of 17 UK platforms lacked effective age verification. DeepSee found 60% of 4,346 companion apps have minor-accessible ratings. If you're building in this space, compliance is a cost center, not an afterthought.

Data Privacy

The privacy record is poor.

• Ada Lovelace Institute: 14 of 16 UK platforms claim rights to train on chat logs.
• Mozilla 2024: 10 of 11 romantic companions failed minimum security standards.
• Muah.ai leaked 1.9 million emails with explicit prompts, according to Have I Been Pwned.
• Chattee Chat exposed 43 million messages, per Cybernews 2025.
• InsightAce Analytic: 75% of consumers hesitate to share personal data.

If you're a user, assume your chats are training data unless the policy says otherwise. If you're an operator, assume a breach is a when, not an if.

Outlook

The market is real and growing. The size depends on what you count. Consumer companions likely earn hundreds of millions to low billions a year, not tens of billions. Growth is strongest in paid memory, voice, and media. Regulation and privacy are the key risks.

For a broader comparison of specific apps by price, memory, and privacy, see our OurDream AI alternatives roundup.

Bottom Line

If you're an investor, use the narrow numbers. The broad ones include too much. If you're a founder, know that the top 10% of apps take 89% of spending, and churn is brutal. If you're a user, read the privacy policy before you tell a chatbot anything you wouldn't put in an email.

The AI companion app market is not a bubble. It's also not as big as the biggest headline. Somewhere in the middle is the truth, and it's still growing.