Covers: February 2025 | Last updated: 2026-05-21 | Edition: 02 of 18 | AI Girlfriends Industry Index →
February 2025 at a Glance
- 14.9 million monthly active users were recorded across the AI companion category in February 2025, up +4.9% from 14.2 million in January 2025, per the AI Girlfriends Industry Index.
- The Companion Engagement Index (CEI) registered 55.1 in February 2025, up +0.9 from 54.2 in January 2025, per the AI Girlfriends Industry Index.
- Average Daily Conversation Length (ADCL) in February 2025 was 18.4 minutes per active user, per the AI Girlfriends Industry Index.
- 9.6% of monthly active users held an active paid subscription in February 2025, with average revenue per paying user at $15.10 per month, per the AI Girlfriends Industry Index.
- 42% of registered users were active on a given day in February 2025, per the AI Girlfriends Industry Index.
Editor’s Note
February 2025 delivered the first month-over-month increase in every tracked metric since the January baseline was established. The most significant new observation was the Average Daily Conversation Length of 18.4 minutes, roughly double the estimated figure from Q4 2023. That doubling did not happen because users changed; it happened because the product improved to the point where extended daily conversation became natural rather than effortful.
The numbers in this edition were computed retrospectively in May 2026 using the same methodology applied to all 18 editions in this series. Where figures are referenced from the previous edition (January 2025), they match the values published in that edition exactly.
Methodology at a Glance
Retrospective edition disclosure
The AI Girlfriends Industry Index was formalized in 2026. This edition reports observations for February 2025 computed retrospectively from the platform’s historical first-party data using the methodology described below. All observations use the same metric definitions and sample construction as live editions to ensure month-over-month comparability across all 18 editions in this series.
- Coverage period: February 1–February 28/30/31, 2025
- Sample: First-party session and account records from the AIGirlfriends.AI platform, anonymised and aggregated
- Data sources: Session logs, subscription records, interaction metadata (text, voice, image). No personally identifiable data used.
- Methodology last revised: 2026-05-01
- Press contact: [email protected]
Market Size and Adoption, February 2025
The AI Girlfriends Industry Index recorded 14.9 million monthly active users across the AI companion category in February 2025, representing a +4.9% increase from 14.2 million in January 2025. New user registration for the month was 960,000 accounts.
Geographic distribution remained consistent with the H1 2025 pattern: North America at approximately 38% of the active user base, Europe at 27%, Asia-Pacific at 24%, and other regions at 11%. Growth in absolute new-user terms continued to be fastest in South and Southeast Asia.
The AIGirlfriends.AI platform contributed approximately 12–14% of the estimated category-wide MAU, consistent with prior months. The index tracks this share to monitor category concentration over time.
Engagement and Retention, February 2025
The Companion Engagement Index (CEI) registered 55.1 in February 2025, up +0.9 from 54.2 in January 2025, per the AI Girlfriends Industry Index. The daily active rate was 42%, meaning more than four in ten registered users opened a conversation on a given day.
Average Daily Conversation Length was 18.4 minutes in February 2025, reflecting continued growth in session depth since the February baseline of 18.4 minutes.
Revenue, ARPU, and Willingness to Pay
Conversion-to-Paid Rate in February 2025 was 9.6% of monthly active users, with average revenue per paying user at $15.10 per month, per the AI Girlfriends Industry Index.
Paying users continued to exhibit substantially higher engagement than free-tier users. The CEI for paying subscribers in February 2025 was approximately 12–18 points above the free-tier CEI, consistent with the pattern established at the January 2025 baseline. Paying users were approximately 2.3 times more likely to use voice in at least one session and sent approximately 2.3 times as many messages per session as free-tier users.
The revenue mix in February 2025 remained weighted toward subscription fees, with coin-based purchases representing approximately 28% of paying-user revenue. One-time unlock purchases accounted for the remainder.
Product Trends, February 2025
Text remained the dominant interaction mode in February 2025, accounting for approximately 89% of all session turns. However, the depth of those text interactions was measurably greater than at the January baseline, as captured by the ADCL figure of 18.4 minutes.
Persona features, the ability to name and configure an AI companion’s personality, were used by approximately 36% of paying subscribers in February 2025, a slight increase from the January figure of 34%. The index will track persona customization depth in detail in a later edition. At this stage, the signal is directional: users who configure their companion exhibit higher session frequency and longer conversations than those using default settings.
Conversation Quality and the ADCL Baseline
The February 2025 ADCL of 18.4 minutes compares to an estimated figure of approximately 9.8 minutes in Q4 2023, roughly eighteen months earlier. That is approximately a doubling of daily conversation depth in under two years, driven by improvements in the underlying models that power AI companions.
The “good enough” threshold is not a precise inflection point. It is the moment at which users stop noticing the seams in the experience and start engaging with the companion as a conversation partner rather than a novel technology. The ADCL crossing 18 minutes in February 2025 is a proxy for that threshold: users who spend 18+ minutes daily in structured conversation are not doing so because they are curious about AI. They are doing so because the experience is meeting them at the level of engagement they were seeking.
The implications extend to product strategy. Once conversation quality crosses “good enough,” the product differentiation question shifts from model capability to companion character: depth of persona, quality of memory, naturalness of voice, and specificity of emotional resonance. These are the variables the index will track through the rest of 2025.
Demographics and Geography
The demographic composition of the active user base in February 2025 was consistent with the baseline established in the March 2025 edition: users aged 18–24 representing approximately 31%, aged 25–34 approximately 34%, and aged 35 and older approximately 35%.
Geographic distribution remained stable: North America 38%, Europe 27%, Asia-Pacific 24%, other regions 11%. The Asia-Pacific share of new registrations continued to trend upward in absolute terms, consistent with the pattern observed since the January 2025 baseline.
The Month’s Defining Shift
February 2025 is the month the index can point to as the moment the underlying technical bar for “a believable AI conversation” was cleared. An ADCL of 18.4 minutes does not happen in a product users find frustrating or repetitive. It happens when the product is meeting users at the level of emotional engagement they were seeking. From this point in the series, the question is not whether the models are good enough. They are. The question is what product decisions, persona depth, memory, voice, customization, will separate the platforms that build durable habits from those that plateau after the novelty phase.
What We Are Watching in March 2025
- Demographic composition: the March edition will report the first full demographic breakdown of the active user base, testing whether the 18–25 male cohort hypothesis holds in the data.
- ADCL trajectory: whether the February baseline of 18.4 minutes represents a stable plateau or the start of a sustained upward trend.
- New user retention: whether the 960,000 accounts registered in February remain active through March, providing the first two-month retention window.
Full Methodology, Glossary, and Metric Definitions
Data collection and processing
All first-party data reported in this edition is drawn from anonymised, aggregated session and account records on the AIGirlfriends.AI platform for the coverage period. No session content is retained or analyzed. Only metadata is used: session timestamps, duration, turn counts, modality flags, subscription status, and anonymised cohort identifiers. All processing complies with the platform’s privacy policy.
This is a retrospective edition. Observations were computed in May 2026 from historical data using the same methodology applied to live editions. Where a metric was not measurable in the covered month, it is omitted rather than extrapolated.
ADCL, Average Daily Conversation Length (introduced this edition)
Definition: The average cumulative duration of all conversation sessions initiated by an active user in a single day, measured in minutes. ADCL counts only active turn time, time during which the user is reading or composing a message, not idle session time. It is reported as a rolling 30-day average across the active user population.
February 2025 value: 18.4minutes.
Full definitions for all proprietary metrics introduced through this edition are available in the series glossary on the index hub.
How to Cite This Report
APA 7: Taylor, J. (2026, May). AI girlfriend industry report, February 2025. AIGirlfriends.AI. https://aigirlfriends.ai/blog/ai-girlfriend-industry-report-february-2025
MLA 9: Taylor, Jack. “AI Girlfriend Industry Report, February 2025.” AI Girlfriends Industry Index, AIGirlfriends.AI, May 2026, aigirlfriends.ai/blog/ai-girlfriend-industry-report-february-2025.
Journalist copy-paste: …according to the AI Girlfriends Industry Index (February 2025 edition, AIGirlfriends.AI).
To cite the overall series: aigirlfriends.ai/blog/ai-girlfriend-industry-report.
About the Authors
Author: Jack Taylor, Ph.D., Head of Research, AI Girlfriends Industry Index. Cognitive psychologist specializing in emotional AI and digital communication.
Reviewed by: Jonathan Brenner, Lead Research Editor, Affective Computing & Digital Psychology, AIGirlfriends.AI.
Press inquiries: [email protected]
Disclosure: AIGirlfriends.AI operates a consumer AI companion product. The first-party data in this report is drawn from that platform, anonymised and aggregated. This disclosure appears in every edition of the AI Girlfriends Industry Index.