OpenAI’s ChatGPT finished 2025 as the single most‑downloaded smartphone app worldwide, with TikTok, Instagram, Facebook and WhatsApp rounding out the top five — a ranking that underlines how generative AI and short‑forrm social video reshaped mobile attention and monetization last year.
Background
Sensor Tower’s 2026 “State of Mobile” report — its annual, cross‑platform look at downloads, time‑spent, and in‑app purchase (IAP) revenue for calendar 2025 — frames the story: generative AI became the year’s breakout engine while social media continued to hoard the majority of attention. The firm reports record totals for IAP revenue and aggregate hours spent, and places AI assistants and a new “Short Drama” vertical at the center of 2025’s growth dynamics.
At the same time, independent trackers published alternative tallies that differ in scale and some headline figures. Appfigures’ 2025 accounting, for example, estimates global app spending at about $155.8 billion and records overall downloads materially lower than some Sensor Tower figures — a reminder that methodology (panel compge, and the unit of measurement) matters when comparing vendor reports. Readers should treat absolute download totals with care and focus on the directional trends that are consistent across providers.
What the numbers say: downloads, time spent, and money
The big picture
- Sensor Tower highlights an overall migration of value on mobile from pure install‑chasing to monetization‑first models: more spending per active user, driven by subscriptions, generative AI features, and creator commerce.
- Social apps continued to dominate attention: more than 60% of total time spent across mobile apps in 2025 occurred within social and messaging apps, even as downloads in the social bucket cooled slightly.
- AI assistants became a top growth vector: Sensor Tower and multiple press summaries document explosive increases in downloads, session volume and hours spent within generative AI apps during 2025.
Top downloaded apps by category (high‑level)
The cross‑catalog picture Sensor Tower and contemporary press summaries present is clear and unusual: a new mix of entrants (mobile AI assistants) sits beside longstanding attention magnets (TikTok, Meta apps, Netflix), and fast‑growth retail and short‑drama players appear in download lists. Among the notable category leaders reported for 2025: ChatGPT (AI), TikTok (social), Temu (retail downloads), Uber (travel/e‑hailing) and Netflix (video streaming).
Money: non‑game revenue overtakes games
A major structural shift documented across industry trackers is that non‑game IAP revenue exceeded games in 2025. Sensor Tower attributes this to subscriptions and AI‑driven monetization; Appfigures reports a similar directional outcome while posting a different aggregate dollar figure. Practically, developers now prioritize higher average revenue per user (ARPU) and recurring revenue, rather than raw new‑install scale.
Why ChatGPT, why now
Product‑level explanations
- ChatGPT’s mobile client evolved from a curiosity to a habitual tool during 2025: multimodal inputs (voice, images), model upgrades, and an expanding suite of specialized “GPTs” made it a daily utility for writing, debugging, tutoring and creative ideation. Those product changes produced both a spike in downloads and a dramatic lift in session length.
- The app benes‑platform resonance: web traffic, API integrations, and widespread third‑party embedding created many stickiness vectors that translate into mobile retention and spend. Industry telemetry shows ChatGPT’s usage footprint ballooned through 2025.
Measurement nuance: downloads vs active users
“Most‑downloaded” is a headline that conceals measurement choices. Sensor Tower counts unique downloads on a per‑Apple/Google account basis for the year; other trackers report monthly leading positions or cumulative installs on different bases. Those technical differences can shift placements month‑to‑month even if the underlying trend (rapid ChatGPT growth) is consistent. Always check whether a ranking is monthly, quarterly, annual, or cumulative.
The social media paradox: falling downloads, rising attention
What changed in social downloads
Total social app downloads slipped in 2025 relative to 2024 in Sensor Tower’s accounting, yet time‑spent edged up. The effect reflects market maturation: top social apps already penetrate most addressable users in many markets, so growth now takes the form of deeper engagement rather than more installs. This dynamic favors platforms that continuously improve recommendation systems and creator monetization.
Why attention still concentrates
- Short‑form video and creator economies concentrate attention: TikTok’s feed and recommendation engine captured an outsized share of hours and IAP revenue in 2025.
- Social platforms are now attention platforms more than discovery systems; winning features are those that increase session frequency and session depth (creator payouts, commerce hooks, and AI‑assisted creation tools).
AI assistants: the growth wedge and its implications
The scale of the change
Generative AI apps (specifically AI assistants) displayed dramatic year‑over‑year increases in downloads, sessions and time spent in 2025. Sensor Tower’s report and press summaries report session volumes that passed one trillion and sharp increases in hours spent on AI assistants. Different outlets report varied percentage increases — some summaries cite a +240% jump to ~48.1 billion hours, while other coverage cites even larger multiples for individual apps like ChatGPT. Those differences reflect whether the metric is category‑wide hours, a single‑app total, or includes web‑based usage. Treat any single percentage with caution and always check the vendor’s metric definition.
Why AI usage rose so fast
- Utility meets habit: assistants performed repeatable, substitutable tasks (drafting, summarization, coding, translation) that naturally generate many short sessions and high session frequency.
- Monetization levers: AI features made premium tiers more compelling, accelerating subscription conversions and IAP revenue for firms that integrated advanced models into paid plans.
Risks and open issues
- Privacy and data governance: widespread use of cloud‑based LLMs raises data residency, compliance and sensitive information leakage concerns. Enterprises and privacy‑conscious users will push for local‑processing or stricter governance.
- Model safety and moderation: the steep growth in usage increases exposure to hallucinations, biased outputs, and policy failures — areas that will attract regulatory attention and brand risk.
- Concentration risk: a small set of assistant vendors capturing enormous time‑spent could centralize discovery and traffic flows, with downstream commercial consequences.
Retail, tariffs and regional shocks: Temu & SHEIN
Downloads vs usage
Temu and SHEIN ranked highly in retail downloads in 2025, but their momentum showed signs of strain in top markets. Sensor Tower and concurrent coverage point to slowing expansion and the impact of
policy shocks — notably new U.S. tariffs and rising scrutiny of Chinese‑origin discount retail platforms — which limited their growth in key geographies. At the same time, Amazon retained leadership in active usage by MAU.
Structural lessons for retail apps
- Discounts and UA spend can drive downloads, but long‑term value depends on retention, logistics, and localized trust signals (returns, payment options).
- Tariffs and geopolitical friction are tangible product risks for global retail apps; some growth strategies will need to account for trade policy volatility.
Video streaming and the short‑drama insurgency
Why streaming downloads surged
Video streaming saw a year of revival: downloads for streaming apps rose strongly due to a wave of
short drama formats and mobile‑first episodic experiences that blurred the line between social short‑form and traditional OTT. Sensor Tower reports sizable increases in downloads and time‑spent for streaming apps, with concomitant growth in IAP revenue for in‑app transactions.
Implications for Windows users and PC ecosystems
Short‑drama and vertical video trends primarily affect mobile consumption patterns, but they create desktop implications as well: streaming services will press for cross‑device continuity, richer web players, and Windows‑native apps or PWAs that sync watch history and subscriptions. Developers should prioritize seamless cross‑device sessions to capture users who move between phone and PC during a binge.
Travel, sports and otherowth, new frictions
- Travel/e‑hailing apps grew in downloads and time‑spent, with Uber retaining top ranking. Growth was strongest in emerging markets where mobile adoption and travel infrastructure are both expanding.
- Sports app downloads rose to record levels as fandom and live sports consumption increased, but IAP revenue in sports fell modestly — a signal that conversion within sports streaming remains fragile, partly due to piracy challenges. Authorities and platforms continued crackdowns on piracy outlets, which reshaped distribution and monetization dynamics for sports broadcasters.
Measurement, methodology and why the numbers differ
Major vendors report different totals
Sensor Tower’s State of Mobile and Appfigures’ annual accounting both describe the same underlying market but use different data sources, coverage assumptions, and modeling techniques. The practical consequence is that headline totals — total global downloads, tos spent — will vary between vendors, even while directional conclusions (AI growth, social attention concentration, monetization‑first economics) align. Readers should compare methodology notes before equating the numbers.
Key methodological points to watch
- Unit of measure: daily/monthly active users vs unique annual downloads vs cumulative installs.
- Counting rules: per‑account deduplication (one download per Apple/Google account) vs raw installation events.
- Store coverage: whether third‑party Android markets are included in aggregated totals.
- Time windows and seasonality: small timing differences (e.g., reporting cutoffs) can materially change top‑10 rankings for fast‑moving apps.
If a specific figure is strategically important (for purchasing decisions, developer forecasts, or financial models), cross‑check at least two trackers and request the vendor’s raw or disaggregated view where possible.
Critical analysis: strengths, risks, and what to watch in 2026
Strengths and opportunity vectors
- Generative AI as a utility: AI assistants converted novelty into habit by solving repeatable tasks that create constant session demand; this is a durable product wedge if safety, latency and cost structures are managed.
- Monetization optimization: the move to subscriptions and higher ARPU gives developers better revenue predictability and increases the financial viability of non‑game apps.
- Cross‑platform leverage: apps that synchronize mobile, web and desktop contexts (e.g., ChatGPT, Netflix) extract more lifetime value and build platform defensibility.
Risks and structurrement opacity: inconsistent vendor methodologies create noise for strategists and investors; firms and policymakers must demand clearer definitions to have productive conversations about market power and revenue splits.
- Regulatory scrutiny: model safety, data protection and antitrust pressures are likely to intensify around the handful of platforms that capture most attention and traffic referrals. Lawsuits and policy friction (illustrated by high‑profile industry disputes) could re‑shape distribution dynamics.
- Geography and geopolitics: U.S. tariffs, export controls, and national policy moves can rapidly alter global user acquisition economics for cross‑border players — a tail risk that retail and social apps must plan for.
- Piracy and content leakage: sports and streaming ecosystems face continued revenue erosion risk from illicit distribution, and removal of piracy hubs will intermittently alter consumption patterns.
Practical takeaways for developers, product managers and Windows‑centric readers
For app developers and product leads
- Prioritize monetizatioher than raw installs: invest in subscription UX and recurring value features.
- Treat AI as a product hook only if it solves repeatable workflows — measure session frequency and retention tightly.
- Invest in cross‑device continuity (mobile → web → Windows) to capture and retain users who shift contexts.
For security and privacy teams
- Prepare for increased scrutiny of data flows to cloud LLMs; adopt encryption, governance and on‑device processing where practical.
- Audit inputs and model outputs for sensitive content leakage, and set conservative defaults for enterprise deployments.
For Windows users and IT pros
- Expect more mobile‑first AI features to appear as Windows integrations (Copilot, native assistant links, and web→app handoff). Plan governance and SSO strategies to manage mobile‑originated AI sessions in corporate contexts.
- Use Progressive Web Apps (PWAs) and official desktop clients where available to preserve session state and reduce friction between phone and PC.
Verification notes and cautionary flags
- Where Sensor Tower reports an IAP total of about $167 billion for 2025, Appfigures’ estimate is lower at roughly $155.8 billion. Both vendors agree on direction (record non‑game revenue growth) but differ on the exact dollar figure; this is a measurement divergence, not contradiction on trend. Treat the dollar totals as vendor estimates rather than absolute ground truth.
- The percent increases reported for AI time‑spent vary by outlet and metric definition. Some summaries cite an approximate +240% growth to 48.1 billion hours for AI apps, while other press pieces highlight even larger multiples for ChatGPT specifically. These differences come from whether the metric includes web sessions, how hours are allocated to app categories, and whether single‑app or entire‑category comparisons are used. Any single percentage should be read alongside the vendor’s methodological note.
- When a headline asserts “most downloaded” or “largest” be sure to verify the reporting window and unit (monthly vs annual, unique accounts vs total installs). Ranking volatility can be high for fast‑growing apps; month‑by‑month leaders will change even if annual leaders are stable.
What to expect in 2026: scenarios and signals to monitor
- Signal: continued migration of advertising and commerce spend into platforms that also host AI assistants. Scenario: AI‑driven referral traffic grows, changing how brands value discovery channels.
- Signal: consolidation among assistant vendors around enterprise integrations and regulated data zones. Scenario: enterprise deals and compliance milestones become gating factors for consumer growth in certain markets.
- Signal: regional policy shifts (tariffs, content regulation) that alter global user acquisition economics. Scenario: some retail and entertainment players refocus on regionally resilient markets and diversify supply chains.
Conclusion
The mobile market that closed out 2025 is recognizably different from the one that opened 2023: generative AI moved from novelty to habit, short‑form and short‑drama formats rewired how video is discovered, and monetization strategies shifted decisively toward recurring revenue. Sensor Tower’s State of Mobile 2026 places ChatGPT, TikTok and other dominant apps at the center of this transition, while independent trackers and news outlets confirm the broader directional shifts even as absolute totals differ between vendors. Developers, platform owners and Windows‑centric IT professionals should prepare for a market where attention is more concentrated, monetization is more strategic, and regulatory and privacy pressures will increasingly shape product decisions.
Source: technext24.com
ChatGPT, Temu, Uber, and Netflix top most downloaded apps by global smartphone users in 2025 - Technext