India’s mobile app market generated $345 million in consumer spending in the second quarter of 2026, up 35% from a year earlier, according to a report by Sensor Tower shared with TechCrunch and published on 31 July 2026. That is a checkable claim, it comes from one analytics firm’s estimate of what Indian consumers paid inside apps, on subscriptions, in-app purchases, and one-time buys, over a three-month window. It is also, on its own, a number that flatters. The more useful question is what changed to produce it, and whether “India is starting to pay for apps” is the right way to describe that change.
I think it is partly right, and partly a story that needs its caveats stated as loudly as its headline. India has been the world’s largest app download market for years and, by reputation and by the data now available, one of the hardest places in the world to turn a download into revenue. Sensor Tower’s Q2 numbers say the revenue side is moving. Appfigures’ download numbers, covering the same market over a longer window, say the download side has not moved at all since 2023. Both things being true at once is the actual story, and it is a narrower, more specific claim than “Indians are paying for apps now.” Indians who were already using apps are paying somewhat more. That is worth writing about carefully, not celebrating.
A Market That Downloads Everything and Paid For Almost Nothing
The starting point, and the reason this quarter’s number is being written about at all, is India’s long-standing reputation as an app market that generates enormous download volume and comparatively little revenue. Appfigures puts India’s quarterly app download volume at 6.3 billion, a figure that has stayed broadly stable since 2023. That is an extraordinary number in absolute terms. It also means India’s download base is not growing, at least not in a way that shows up in Appfigures’ data over the past three years. Whatever is producing the 35% jump in spending that Sensor Tower is reporting, it is not new people picking up their phones and installing apps for the first time.
That distinction matters more than it might look at first glance. A market where downloads are rising and revenue is rising too is a market getting bigger. A market where downloads are flat and revenue is rising is a market changing in composition, in behavior, or in price, not in size. India, on this evidence, is the second kind of market right now. The 700-plus million smartphone users who already have a phone full of apps are, in aggregate, spending more inside the apps they already have. New downloads are not driving the growth. Existing downloads are converting into paying relationships at a higher rate than before, or the same relationships are paying more per user, or both. Sensor Tower and Appfigures do not, between them, tell us which.
The Arithmetic, Done in the Open
Here is where I want to be transparent about a calculation I am making myself, because it is not a number either Sensor Tower or Appfigures published, and I do not want to present it as though it were.
Sensor Tower reports $345 million in India app consumer spending for Q2 2026. Appfigures reports 6.3 billion app downloads in India per quarter, a figure that has held roughly steady since 2023. Divide one by the other; $345 million across 6.3 billion downloads works out to a little under 5.5 cents of consumer spending per download for the quarter.
Set that next to what Sensor Tower reports as revenue per download in three other markets: $4.60 in the United States, $3.90 in South Korea, and $6.10 in Japan. On my own arithmetic, the gap between India and the United States is somewhere in the neighborhood of 80 times.
I want to state the limits of that comparison plainly, because it is the single most attention-grabbing number in this piece and it is also the one I built myself rather than the one Sensor Tower handed me. Sensor Tower’s $4.60, $3.90, and $6.10 figures for the US, South Korea, and Japan are presented as “revenue per download,” and it is not stated in the material available to me whether that figure is calculated the same way I calculated India’s, by dividing total quarterly consumer spending by total quarterly downloads for that market, over the same quarter, using the same download-counting methodology. It could be a lifetime value figure, a different time window, or a metric computed on a different download base (first-time installs only, say, versus all installs including reinstalls). If Sensor Tower’s own per-download figures for those three markets are not simple division of quarterly spend by quarterly downloads, then my 5.5-cent India figure and their $4.60 US figure are not strictly the same measurement, and the 80-times gap, while directionally almost certainly real given how far apart the inputs are, should be read as an order-of-magnitude estimate rather than a precise multiple.
What I am confident saying, because it rests on the two headline figures Sensor Tower and Appfigures each published rather than on any blending of methodologies: India generates roughly $345 million in app spending across roughly 6.3 billion downloads in a quarter, a low-single-digit-cents-per-download outcome by construction, while the report’s own figures put the United States at $4.60 per download, South Korea at $3.90, and Japan at $6.10. However the per-download comparison is ultimately reconciled, those are different orders of magnitude, not a rounding difference.
Table: The Numbers Side by Side
| Metric | Figure | Source |
|---|---|---|
| India app consumer spending, Q2 2026 | $345 million | Sensor Tower |
| India Q2 2026 spending growth, year on year | 35% | Sensor Tower |
| India quarterly app downloads (stable since 2023) | 6.3 billion | Appfigures |
| India spend per download, Q2 2026 (my calculation) | approx. 5.5 cents | Derived from Sensor Tower + Appfigures |
| Revenue per download, United States | $4.60 | Sensor Tower |
| Revenue per download, South Korea | $3.90 | Sensor Tower |
| Revenue per download, Japan | $6.10 | Sensor Tower |
| Non-gaming share of app revenue, H1 2026 | 68% | Sensor Tower |
| Non-gaming share of app revenue, three years earlier | 58% | Sensor Tower |
| ChatGPT + Claude share of India AI app revenue, Q2 2026 | 83% | Sensor Tower |
| ChatGPT India daily revenue, currently | approx. $60,000 | Appfigures |
| ChatGPT India daily revenue, October 2025 | approx. $80,000 | Appfigures |
| ChatGPT India downloads, past month | approx. 1.8 million | Appfigures |
| Q2 2026 app revenue growth, India | 35% | Sensor Tower |
| Q2 2026 app revenue growth, Mexico | 30% | Sensor Tower |
| Q2 2026 app revenue growth, Turkey | 25% | Sensor Tower |
| Q2 2026 app revenue growth, United States | -3% | Sensor Tower |
A Note on the $345 Million Versus “More Than $200 Million”
Readers who go looking for the underlying Sensor Tower report may run into a second figure, “more than $200 million in quarterly consumer spending,” used in the same report in the context of comparing India’s growth rate against other major app markets. Having read the source material directly, I could not find evidence that this is a separate spending total distinct from the $345 million headline figure. It reads as a looser, rounded-down restatement of the same Q2 spending number, made in a sentence about India’s growth rate outpacing other markets rather than in a sentence stating the market’s total size. I am treating $345 million as the precise, citable figure for India’s Q2 2026 consumer spending, because that is the number stated plainly and attributed clearly to Sensor Tower as the headline total. Where the report uses “$200 million-plus” in a comparative aside, I am not treating that as a second, independently sourced figure, and I would rather flag the ambiguity than silently pick a number and move on.
Where the Growth Is Actually Coming From: Country Comparison
Sensor Tower’s growth comparison across markets is useful for context, and worth taking at face value rather than folding into the per-download arithmetic above, since it is a rate, not a level, and rates are less sensitive to the base-methodology questions I raised a moment ago. India’s 35% year-on-year growth in Q2 2026 outpaced Mexico’s 30% and Turkey’s 25%, all markets usually grouped together as large, price-sensitive, mobile-first economies. The United States, by contrast, saw app revenue decline 3% in the same quarter, according to Sensor Tower. That is a genuinely interesting data point on its own. It suggests that whatever growth story is playing out in app monetization globally right now, it is concentrated in markets like India, Mexico, and Turkey rather than in the mature US market, where app spending on Sensor Tower’s numbers actually contracted.
That framing supports a version of “India is starting to pay for apps” that is more defensible than the flat version: not that India has caught up to anything, but that India is one of a small cluster of large, historically low-monetization markets where app revenue is now growing meaningfully faster than in the United States, even though the absolute level of spending per download remains far apart. Growing 35% from a small base and shrinking 3% from a huge base are two different stories, and conflating them, saying only “India grew faster than the US,” would flatten a comparison that needs both the rate and the base stated together. I have tried to keep both in view throughout this piece for that reason.
From Games to Everything Else
One of the more concrete shifts in Sensor Tower’s data is the changing composition of what Indians are actually paying for inside apps. Non-gaming apps accounted for 68% of India’s app revenue in the first half of 2026, up from 58% three years earlier, per Sensor Tower. Games have historically been where mobile monetization concentrates in most markets, through in-app purchases, loot boxes, and battle passes aimed at a relatively small share of highly engaged spenders. A ten-percentage-point shift away from games and toward everything else, over three years, is a meaningful reallocation of where the money is coming from.
It does not tell us, on its own, exactly which non-gaming categories are absorbing that share. Sensor Tower’s report, as summarized to me, groups AI apps, entertainment apps, and other premium categories together under the “non-gaming” umbrella, and the 83% concentration figure discussed below suggests AI products are doing a disproportionate amount of the work inside that 68%. What the shift does tell us, reasonably confidently, is that Indian app spending is broadening out from the games-dominated pattern that shaped mobile monetization in most markets for the past decade, toward a mix where subscriptions, tools, and AI assistants make up a much larger share of what gets paid for. That is a different kind of willingness to pay than loot-box spending. Subscriptions imply a recurring judgment that a product is worth a monthly fee, repeated every billing cycle, rather than a single impulse purchase inside a game session.
Two Products, 83% of a Category
If the non-gaming shift is one axis of change, the concentration of AI spending is another, and a much narrower one. OpenAI’s ChatGPT and Anthropic’s Claude together accounted for nearly 83% of India’s AI app revenue in Q2 2026, according to Sensor Tower. That is a striking degree of concentration for a category that, on paper, includes a wide and growing field of AI chatbots, coding assistants, image generators, and productivity tools, many of them backed by well-funded competitors with their own India strategies.
An 83% share held by two products says a few things at once. It says the AI-spending story in India right now is really a ChatGPT-and-Claude story, not a broad AI-adoption story, at least on the revenue side; other AI apps may have meaningful download numbers without meaningfully denting that revenue share. It also says the category is fragile in a specific way: a large swing in either product’s India performance moves the whole category’s numbers, not just that one product’s numbers. Which is exactly what the next section is about.
It is also worth being honest about what “83% of India’s AI app revenue” does not establish. It does not tell us what share of India’s total $345 million in app spending is AI spending at all, only what share of the AI slice belongs to these two products. Nor does it tell us whether that 83% concentration is unusually high for India specifically, or simply reflects that ChatGPT and Claude are the two most prominent consumer AI products globally, in which case a comparably concentrated split might show up in most large markets. Sensor Tower’s report, as summarized, does not include that cross-market comparison, so I am not asserting one.
The Complication: ChatGPT’s India Revenue Is Falling
This is the part of the story that a simple “India is finally paying for apps” headline would prefer not to include, and it deserves the same space as the good news above it.
Appfigures puts ChatGPT’s India revenue at roughly $60,000 a day currently, down from roughly $80,000 a day in October 2025. That is a decline of about a quarter in daily revenue over roughly nine months, for the single product that, together with Claude, accounts for 83% of India’s AI app revenue. And it has happened while downloads kept coming: Appfigures counts approximately 1.8 million ChatGPT downloads in India over the past month, which is not a market in retreat by any obvious measure of interest.
Downloads holding up while revenue falls is the same shape of disconnect, in miniature, as the flat-downloads-rising-spend pattern across the whole Indian app market, except running in the opposite direction for this one product. People keep installing ChatGPT in India. Fewer rupees, or a lower average amount per paying user, are flowing back to OpenAI from that installed base than were nine months ago.
I do not have a sourced explanation for why, and I am not going to invent one. There are several plausible readings, and I want to lay them out as readings, not conclusions. One is pricing: if ChatGPT ran a lower-priced India tier, or a promotional rate, in and around October 2025 that has since lapsed or been adjusted, average revenue per paying user could fall even as the user base holds steady or grows, depending on how the adjustment moved the conversion rate. Another is that October 2025 revenue may have been boosted by a promotional period, a launch, or a seasonal effect (India’s festive season, which includes major shopping and gifting occasions, falls in that window most years) that inflated the baseline the current number is being compared against, making the “decline” partly an artifact of comparing against an unusually strong month rather than a normal one. A third is a shift in who is subscribing: if new signups skew toward free-tier users, or toward lower-priced regional plans, while some earlier paying cohort lapses or downgrades, total daily revenue could fall even with downloads and even total user counts holding up. None of these is stated in the reporting available to me. The report does not explain the decline, and neither can I, beyond naming the shapes an explanation could plausibly take.
What I can say is that this single data point cuts directly against the tidiest version of this story. If India’s AI apps are 83% concentrated in two products, and the larger of those two products has seen its India daily revenue fall by roughly a quarter over nine months, then a meaningful share of the category driving the “India is starting to pay” narrative is currently moving in the wrong direction, even while the aggregate India app spending number for Q2 2026 moved in the right one. Both things are true in the same quarter, from the same set of reports, about overlapping parts of the same market. That is not a contradiction that needs resolving into a single tidy narrative. It is what the data actually shows, and a data-led piece about this topic should say so rather than pick the more flattering half.
What This Means for Developers Building for India
For a developer or a company building software aimed at Indian users, the useful reading of all this is narrower and more specific than “India is warming up to paid apps,” and probably more useful because it is narrower.
First, the growth on the table is coming from existing users, not new ones. Appfigures’ flat download numbers since 2023 mean an India strategy premised on riding a growing download base to growing revenue is building on a foundation that, per this data, has not been growing for three years. Whatever revenue growth is available in the near term looks like it has to come from converting or upselling users who are already there, not from acquiring new ones, at least judging by the download trend this report describes.
Second, the 68% non-gaming share, up from 58%, is a signal about category, not about price tolerance broadly. It says Indian consumers are more willing to pay for something that isn’t a game than they were three years ago, which is consistent with subscription products, AI tools, and premium utilities gaining ground. It is not evidence that Indian consumers have converged on US or Japanese pricing expectations; the roughly 80-times gap in my per-download arithmetic, caveats and all, argues against reading the non-gaming shift as a pricing-power story. A wider set of Indian consumers appears willing to pay something for the right non-gaming product. There is no figure in this reporting suggesting they are willing to pay US-level amounts for it.
Third, the AI concentration and the ChatGPT decline together are a warning against treating “AI monetizes well in India” as a solved formula to copy. Two products hold 83% of a category whose largest member’s India revenue just fell by a quarter in nine months. That is not a stable base rate to build a business plan on; it is closer to a live experiment whose outcome has already shifted once during the period covered by this report.
What the Data Does Not Show
It is worth closing on the limits, because the honest version of this story is narrower than the headline version, and that is the version worth publishing.
Sensor Tower’s $345 million and 35% figures establish that India’s app spending grew, in aggregate, in Q2 2026, and grew faster than three comparison markets Sensor Tower named, while the United States shrank. Appfigures’ 6.3 billion download figure, stable since 2023, establishes that this growth did not come with a comparable download surge, which is why the growth reads as more spending per existing user rather than more users. My own division of those two Sensor Tower and Appfigures figures produces a roughly 5.5-cents-per-download number for India against a reported $4.60 in the United States, a gap of a rough order of 80 times, a comparison whose precision is limited by an unstated methodology question about how Sensor Tower’s own per-download figures for the US, South Korea, and Japan are calculated. The 68% non-gaming share tells us the composition of spending has shifted toward non-gaming categories, without telling us exactly which non-gaming categories, beyond the strong hint given by the 83% AI concentration figure. And the ChatGPT revenue decline tells us that even inside the AI category driving much of this story, the picture is not uniformly upward, for reasons this reporting does not establish.
None of that adds up to “India is finally paying for apps” as a clean, closed story. It adds up to something narrower and, I think, more honest: a large, historically low-monetizing market where aggregate spending is rising faster than in most comparable markets, where the composition of that spending is shifting away from games, where two AI products dominate a fast-growing category, and where the single largest product in that category has, in the same period this growth is being reported, seen its own India revenue fall by roughly a quarter. Whether that adds up to a durable shift in how India pays for software, or a snapshot of one unusually active quarter inside a market that is still, on a per-download basis, worlds apart from the United States, Japan, and South Korea, is not something this data, on its own, can settle. It is worth watching the next two or three quarters of the same figures, from the same two firms, before drawing a firmer conclusion than that.