Chasing 100M Daily Users, Duolingo Reconsiders Ads, Paying Influencers, and Its Most Expensive Plan

This year, Duolingo announced a major strategic shift: from monetization to user growth, with a target of 100 million daily active users by the end of 2028.

To get there, it needs to grow about 25% year over year, every quarter. In Q2 2026, daily users grew 23%, to 58.7 million, the first time growth has sped up in two years.

Duolingo also disclosed a roughly $10 million bonus that pays out if Q4 growth hits 25% or more.

Paid subscribers grew by about 200,000 in the quarter, the fewest since 2021.

I read the earnings call transcript. The target is pushing Duolingo to change its strategy, and in some cases, beliefs it has held for years. Advertising, whose job at Duolingo was to annoy free users into upgrading, is getting a new one.

AI also got so cheap that the feature Max was built around is now in the cheaper Super plan. Max may not survive, and a new ad-supported plan is being tested.

Duolingo Q2 2026: By the Numbers

* Q3 2025 profit includes a one-time $256.7 million tax benefit. Strip that out, and it was in line with the others.

This quarter, Duolingo brought in $298.5 million in revenue, up from $252.3 million a year ago. It also had its smallest profit since the end of 2024: $33.2 million, down from $44.8 million a year ago and $43.5 million last quarter.

But make no mistake, Duolingo is still a cash machine. It made $78.6 million in cash in the quarter, expects more than $375 million this year, and is sitting on $1.3 billion with no debt. It also raised its profit target for the year, from the 25% of revenue it set in February to 26.5%.

The gap between the falling profit and the healthy cash is mostly accounting. Subscriptions are about 86% of Duolingo’s revenue, and if you buy a yearly one today, Duolingo pockets all of your money now but only counts it as revenue a bit at a time over the next year.

That upfront cash shows up as bookings, which exceeded revenue for over three years. But in Q2 2026, bookings fell to $289.1 million against $298.5 million in revenue. In fact, they have slid two quarters in a row.

Duolingo is trading money for users.

How Duolingo Got Its Growth Back

To get to 100 million, Duolingo needs to make existing users stickier and find new ones.

It started by making the app less annoying. Fewer ads, which we noticed earlier this year, and free trials going from seven days to a month. Co-founder and CEO Luis von Ahn was blunt about why: “We were doing a number of things that were counter to DAU growth. So we kind of had a little bit of a reset.”

In June, Duolingo let learners restore their longest-ever streak by completing three lessons. It was a huge success. 15.4 million took it. Nearly 8 million of them had no streak going at all.

That one was a one-off, and Duolingo says so. But it also ships a new version of the app every week with around 350 changes in it, keeping whatever works. It calls this the “Green Machine.” No single change does much. Von Ahn, on what drove the quarter: “it’s hard to point to something that actually did it.”

All of this works on people Duolingo already has, and that won’t deliver 25% growth every quarter. So you go where the people are.

India, China, and Indonesia are three of the four most populous countries on earth. Duolingo is paying hundreds of creators there, and two-thirds of its social media impressions in those countries now come from them rather than from Duolingo itself.

This is new. For years, Duolingo did not pay for growth. It was “100% organic through word of mouth” at first, then the unhinged owl on social media, still clearing a billion impressions a quarter. Now it is doing performance marketing too, something von Ahn says he was once “massively allergic” to.

China is already Duolingo’s second-largest market by daily users, and von Ahn expects it to become the largest in a year or two.

Growth in monthly users was under 6% at the start of the year. It is back to 10%.

Wrong About Math

Growth is not coming from Duolingo’s newer subjects. Math and music have single-digit millions of daily users each, well behind chess. Von Ahn: “We have 60-some million daily active users. Even a 50% growth here doesn’t contribute all that much.”

Asked about math specifically, he was blunter than he needed to be:

“Originally when we launched math, I thought, and I was wrong, I thought that we could get the average person on the street to get addicted to learning math. I was wrong. I don’t believe we can do that now. Our strategy is actually to mainly teach math to the people who actually need to learn math, which are basically K through 12 students.”

The alternative, as he put it: “trying to get your average 35-year-old person who hates math to suddenly love it. I would love to do that, but I’ve given up on that one.”

Duolingo’s whole premise is that gamification can get anyone to learn anything voluntarily. Math is where its founder now says it cannot.

That also leaves the company in an odd spot. It wants K-12 learners but not K-12 institutions: “We may not sell to schools. That’s not our goal.” Three months ago we reported that Duolingo is shutting down Duolingo for Schools, its only classroom tool, in 2027.

Lite Comes In, Max May Be on Its Way Out

Duolingo ended the quarter with 12.7 million paid subscribers, up 17% in a year. That’s its slowest growth as a public company.

It’s also changing what the plans include and what they cost. The main reason is the falling cost of Video Call, the feature Max was built to sell.

Video Call is live speaking practice. When it launched, it cost Duolingo about 30 cents a call. “This is why we decided to put it behind our most expensive plan,” von Ahn said. I wrote about Max at the time: one of six OpenAI launch partners for GPT-4, personalized AI tutors, but at a premium.

It now costs under a cent, with “no loss in quality,” because Duolingo switched to open-source models. More on that in a future article.

Video Call has moved down to Super. Von Ahn: “Most new Super subscribers get video call… we expect that over the next few months we’re going to give video call to existing user subscribers.”

So what is Max for now? Von Ahn does not know either. “What are we doing with Max?” he asked on the call. “I don’t know yet.”

He laid out three options. Super gets a limited number of calls, and Max keeps unlimited. Max gets some new feature. Or, in his words, “we may actually sunset Max.” A decision is coming in the next couple of quarters, and he says they will do it “without a loss of revenue.”

Duolingo is also testing a new ad-supported plan called Lite. It comes with double the Energy, which also refills faster. Energy replaced Hearts in April 2025: Hearts drained when you got an answer wrong; Energy drains even when you get it right. Von Ahn says the idea is to sell it to people who were never going to buy Super: “if we’re pretty sure that you’re not going to buy Super, then we should try to sell you Super Lite.”

In my own testing, the naming was all over the place. The upgrade screen offered Super Family, Super, and Lite (with ads). Max was missing. On the checkout screen, the same three plans were called Family, Individual, and Lite.

Lite was $35.99 a year. Individual and Family were both $119.99. Duolingo’s own pitch for the family plan: “6 accounts for the price of 1.” It also had a back-to-school sale running, 80% off.

This is probably one of the Green Machine’s experiments rather than final pricing. Duolingo tests this stuff constantly, so you might see something different.

Advertising Gets a New Job

If I had to guess, as subscriptions slow, advertising is where Duolingo goes looking for growth. Not ads shown more often, which a company chasing 100 million daily users cannot afford, but ads that pay better, sold directly to brands instead of through the ad networks.

Before subscriptions existed, ads were the business. Duolingo Plus launched in 2017, sold mainly as a way to make them go away.

For years after that, Duolingo did not sell the ads itself. Ad networks did, and the company focused on subscriptions, which now bring in twelve times what ads do. Two years ago, the entire ads operation was “one person whose half-time job was to deal with ads.”

Now there is a team, run by Andrew Guendjoian, who came from BuzzFeed. Duolingo Ads launched last October. Its first brands: Adobe and Marriott Bonvoy.

Von Ahn thinks there is room: “Pretty much every app our size or larger than us makes a ton more money from ads than we do.”

In June, it started testing something different. Watch a 30-second video and get 30 minutes of unlimited Energy. Etsy and Universal Pictures are the first sponsors. The ad gives you more of the app instead of less.

So far, none of this has shown up in advertising revenue. It made $21.1 million this quarter and has been basically flat for four straight quarters.

Duolingo has a little over two years to reach 100 million daily users. More changes are coming, even to things Duolingo spent years refusing to do. Class Central will read the earnings calls, test the app, and connect the dots.

The post Chasing 100M Daily Users, Duolingo Reconsiders Ads, Paying Influencers, and Its Most Expensive Plan appeared first on The Report by Class Central.

​