How's Duolingo doing these days?

Is Duolingo still growing now?

Duolingo is still growing, but it’s more about engagement depth than easy audience expansion.

The latest hard signal came from Duolingo’s May 2026 Q1 shareholder letter. Daily active users reached 56.5 million, up 21% year over year. Paid subscribers reached 12.5 million, also up 21%. Revenue grew 27% to $292 million. Those are clearly not weak numbers for a consumer subscription app at this scale.

But the more interesting signal is underneath the headline. In Q4 2025, Duolingo had 52.7 million DAUs, up 30% year over year. One quarter later, DAUs were still growing, but the year-over-year rate had stepped down to 21%. Management had already warned in February 2026 that DAU growth decelerated through 2025 after never falling below 40% year over year from Q2 2022 to Q2 2025.

So the answer is yes, Duolingo is still growing today. But the free hypergrowth phase is over. The current question is whether Duolingo can turn better learning outcomes, AI speaking practice, and new subjects into the next growth curve.

Is Duolingo sacrificing money these days?

Yes. Duolingo is deliberately giving up near-term money to rebuild user growth.

This is one of the clearest recent signals because the company quantified the tradeoff. In its February 2026 shareholder letter, Duolingo said it was reinvesting more than $50 million of foregone bookings, equal to roughly 5 points of bookings growth, into improving the free-user experience. Management also said 2026 bookings growth would be around 11%, compared with nearly 20% if the company kept operating like before.

That showed up immediately in the numbers. Q1 2026 revenue grew 27%, but bookings grew only 14%. Q2 bookings guidance was even more telling at roughly 5.8% growth, while full-year 2026 bookings guidance sits around 10.5%. For a company that generated 33% bookings growth in FY 2025, this is a real strategic slowdown, not just normal quarterly noise.

The interpretation matters. Duolingo is not being forced into lower monetization because users are leaving en masse. It is choosing to monetize less aggressively because management thinks the bigger long-term risk is product exhaustion. That is a more subtle but more important signal than the headline slowdown.

If you want more recent data on this point, please see our latest EdTech market report.

Is Duolingo’s free product getting repaired now?

Yes. Duolingo has clearly understood that the free experience became too monetized.

The first signal is management’s own language. In February 2026, Duolingo said part of the 2025 DAU slowdown came from an increased focus on monetization. That is unusually direct. Companies rarely admit that growth friction came from their own monetization choices unless the internal data is strong enough.

The second signal is product behavior. Duolingo started moving AI-powered learning value back into the free experience, including more speaking practice for free users, spoken tokens, flashcards, and Speaking Adventures. Class Central also noted that Duolingo made Explain My Answer free again in January 2026 after previously putting it behind Max.

The third signal is user backlash. The Energy system, which replaced Hearts for many users, created visible complaints because it changed the psychology of free progress. Under Hearts, strong users could continue if they avoided mistakes. Under Energy, users felt progress could be throttled even when they were performing well.

Our read is that Duolingo’s 2026 reset is partly defensive. The company is trying to remove enough friction to keep the habit loop alive before AI-native alternatives make switching easier.

Is Duolingo’s AI strategy actually producing results now?

Yes. Duolingo’s AI strategy is producing real operating leverage, especially in content creation.

First, there is the Q1 2026 shareholder letter. Duolingo published 20,500 course units in Q1 2026, versus 7,100 per quarter in 2025 and 1,800 per quarter in 2024. That is more than 11 times the 2024 quarterly output. This is exactly the kind of number that separates “AI slideware” from actual process change.

There is also course depth. Duolingo expanded its 9 most-learned language courses to Duolingo Score 129, roughly CEFR B2. That matters because Duolingo has often been criticized for being strong at habit-building but weaker for advanced, work-ready proficiency. B2-level coverage does not automatically solve fluency, but it moves the product into a more serious learning zone.

Also, current Duolingo hiring pages show open roles tied to AI, machine learning, LLMs, new subjects, scaling operations, monetization, and DET. That tells us the AI push is still visible in how the company is allocating people.

The caveat is cost. Duolingo’s Q1 2026 letter says gross margin should decline toward roughly 69% by Q4 2026 as AI-powered features get more usage. So AI is helping Duolingo move faster, but the bill is moving from model-building into usage economics.

If you want more recent data on this point, please see our latest EdTech market report.

Did Duolingo walk back its AI-first culture recently?

Yes. Duolingo walked back the performative part of AI-first, while keeping the operating strategy intact.

The signal came in April 2026, when Luis von Ahn said on the Silicon Valley Girl podcast that Duolingo had backtracked on evaluating employees based on AI use in performance reviews. Reports from Business Insider and TechRadar both framed the reversal the same way: employees were asking whether they were supposed to use AI even when it did not make sense.

That is important because the original 2025 AI-first memo created backlash around contractors, job replacement, and product quality. The memo had been shared on LinkedIn and said Duolingo would gradually stop using contractors for work that AI can handle, look for AI use in hiring, and make AI use part of performance reviews.

The new signal is more mature. Von Ahn abandoned AI usage as a performative employee KPI. That is probably healthy. A company forcing AI use everywhere often creates theater. A company using AI where it changes content output, speaking practice, experimentation, and personalization has a better chance of compounding.

So Duolingo’s AI culture is less loud now, but not less serious.

Is Duolingo finally serious about speaking practice now?

Yes. Speaking has become the center of Duolingo’s product response to AI tutors.

This was the most important product theme in the May 2026 shareholder letter. Duolingo said speaking practice was historically its biggest gap and is now a core part of the product. Video Call more than doubled the average number of words spoken per user over the past year, and access is being expanded from Max to Super subscribers.

The second signal came earlier, in February 2026, when Duolingo said moving Video Call into Super would give access to about 10 times as many learners. That is a big product philosophy change. Instead of keeping conversation practice as a narrow premium feature, Duolingo is making it a broader retention and learning-outcome feature.

The third signal is feature density. Spoken tokens turn many normal exercises into voice exercises. Flashcards push users to say words out loud. Speaking Adventures create character-led real-world tasks. Together, these features change the product from “tap the correct word” toward “produce language.”

This is Duolingo’s most strategically important move right now. AI-native competitors are trying to own conversation practice. Duolingo is trying to pull that use case back inside its habit loop.

If you want more recent data on this point, please see our latest EdTech market report.

Is Duolingo Max still the main AI bet?

No. The bigger signal is that Duolingo is pushing AI value down into Super and free usage.

Duolingo Max mattered because it proved consumers would pay for AI-assisted language learning. But the recent move is broader. Video Call with Lily is being expanded into Super, free users are getting more speaking practice, and Duolingo is experimenting with changes to Max itself.

That tells us Duolingo has probably learned that AI cannot stay only in the expensive tier. If AI conversation is locked behind Max, users can compare it directly with ChatGPT, Speak, Promova, Praktika, or Google Translate’s AI practice features. If AI becomes part of the broader Duolingo habit loop, replacement becomes harder.

This also explains the margin guidance. Expanding AI features into cheaper or free tiers increases usage cost, which is why gross margin is expected to move down toward 69% by Q4 2026. That is not a random cost issue but the financial footprint of Duolingo choosing defensibility over near-term margin purity.

Max is no longer the whole AI story. Duolingo’s real AI bet is making the core app harder to leave.

Are AI tutors actually threatening Duolingo now?

Yes. AI tutors are a real threat now, especially around spoken fluency.

Speak is the clearest signal. The OpenAI-backed language-learning company reportedly passed $100 million in annualized revenue, reached 15 million downloads, and was valued at $1 billion. Its positioning is very different from Duolingo’s old strength: it is voice-first, conversation-first, and explicitly built around spoken fluency.

Also, Google started adding AI language-learning features inside Translate, including personalized lessons, speaking/listening practice, and later pronunciation practice. Google Translate reportedly helps more than a billion people and handles an estimated trillion words per month, so even a small learning layer inside Translate creates a massive discovery surface.

Recent reviews position Promova as more conversational and real-world than Duolingo, with AI role-play and live tutors. Sensor Tower’s public Promova page showed much smaller scale than Duolingo, but the comparison matters because user expectations are shifting from “daily lesson completed” toward “can I actually speak?”

Duolingo still owns the habit, distribution, and brand. The threat is that AI tutors are attacking the exact weakness Duolingo itself now admits: speaking.

If you want more recent data on this point, please see our latest EdTech market report.

Is Duolingo still dominant in the app stores today?

Yes. Duolingo still has a very strong distribution moat.

Duolingo’s Q1 2026 shareholder letter says its flagship app is the top-grossing Education app on both Google Play and the Apple App Store. That is a major signal because consumer education apps usually struggle to combine large free usage, paid conversion, and daily retention.

Business of Apps’ May 2026 language-learning market update also reinforces the scale gap. It estimated that language-learning apps generated $1.54 billion in 2025, while Duolingo reported $1.03 billion in 2025 revenue. The exact comparability is not perfect because company revenue and app-market estimates are not always measured the same way, but the order of magnitude is still useful: Duolingo is not one player among equals in consumer language apps.

Sensor Tower’s public pages recently showed much smaller estimated monthly revenue for Promova than Duolingo. Even if these third-party app estimates are imperfect, they support the same pattern: AI alternatives are rising, but Duolingo still has the consumer funnel.

That is why the disruption story needs nuance. AI competitors can be better at a specific job, especially conversation. Duolingo is still much harder to beat as a daily consumer platform.

Is Duolingo’s backlash actually hurting the business now?

The backlash is real, but we do not see evidence of structural business damage yet.

There are three separate backlash zones. First, the AI-first memo created concern about contractors, jobs, and quality. Second, the Energy system created user frustration around free progress. Third, the softer 2026 guidance triggered investor pressure, analyst debate, and shareholder-investigation press releases from law firms after the stock drop.

But the operating data does not show a consumer collapse. As seen above, in Q1 2026, DAUs grew 21%, paid subscribers grew 21%, revenue grew 27%, adjusted EBITDA margin was 28.6%, and free cash flow margin was 50.6%. Those are unusually strong numbers for a company supposedly damaged by brand backlash.

The better interpretation is that backlash has changed management behavior before it broke the business. Duolingo is making free usage less painful, widening AI access, softening internal AI mandates, and still buying back shares to offset dilution.

The investor-law-firm activity is worth watching, but right now it looks more like a stock-drop aftershock than proof of operating deterioration.

If you want more recent data on this point, please see our latest EdTech market report.

Is Duolingo becoming more than a language app now?

Yes. Duolingo is increasingly trying to become a consumer learning platform, and Chess is the first serious proof.

The February 2026 shareholder letter said Chess reached more than 7 million DAUs in less than a year, making it Duolingo’s fastest-growing subject. That number is more meaningful than a normal feature launch because it suggests Duolingo’s habit mechanics can travel outside language learning.

The hiring signal points in the same direction. Duolingo’s open roles recently included a Senior Director of Engineering for New Subjects, brand campaign roles, influencer and community leadership, and product roles tied to learning and DET. That looks like a company allocating around multiple growth surfaces, not just maintaining one language app.

Analysts have picked up on this too. BofA’s early-2026 bullish framing reportedly argued that Wall Street was undervaluing Duolingo by treating it only as an education app rather than as a mobile gaming-style platform with subscriptions.

We would be careful not to overstate Math and Music yet. Chess has the strongest proof. But Duolingo is clearly testing whether its real asset is not language content, but habit formation around learning.

Is Duolingo English Test still a real upside?

Yes, but Duolingo English Test looks more like a steady strategic asset than the biggest near-term growth driver.

The acceptance footprint is large. Duolingo English Test lists thousands of accepting institutions globally, including large counts across North America, Europe, Asia, and Oceania. The hiring signal also matters: Duolingo recently had open roles specifically tied to DET, including B2B marketing for the UK, Europe, and ANZ, and senior product roles for DET.

But the growth signal is not clean enough to call DET a breakout engine right now. A TOEFL Resources analysis estimated Q1 2026 DET revenue at $11.3 million and roughly 162,000 tests, implying a decline in volume versus Q1 2025. That is a third-party estimate, so we should not treat it like company-reported truth, but it is still a useful warning signal.

There is also regulatory nuance. UK universities may accept DET for admissions, but UK visa or migration approval is a separate question. That limits how directly institutional acceptance turns into unavoidable test demand.

So DET still matters. It gives Duolingo a credible credentialing business next to the consumer app. But the recent signals do not justify making it the core 2026 growth story.

Did Duolingo’s leadership or finance setup change recently?

Yes. Duolingo changed CFOs right as the story became harder to explain to investors.

In January 2026, Duolingo announced that Gillian Munson would become CFO effective February 23, 2026. This was not an outsider parachuted in after chaos. Munson was already a board member and audit committee chair, while outgoing CFO Matt Skaruppa stayed through the transition after nearly six years at the company.

The timing is the important part. The new CFO inherits a more complex narrative: slower bookings growth, lower 2026 profitability, wider AI feature access, margin pressure from usage costs, a $400 million buyback, and the need to convince investors that 100 million DAUs by 2028 is worth the tradeoff.

This is not a distress signal. The balance sheet still looks strong, with more than $1 billion in cash and no debt in the latest quarterly materials. But it does mark a shift from the simpler “Duolingo keeps beating and raising” era into a more judgment-heavy capital-allocation story.

So, how is Duolingo doing these days?

Duolingo is doing well, but it is in a deliberately uncomfortable transition.

The company is not broken. The latest numbers are still strong: user growth, subscriber growth, revenue growth, free cash flow, app-store leadership, and AI-driven content velocity all point in the right direction. The evidence is strong enough to reject the simple “Duolingo is being killed by AI” narrative.

But Duolingo is also not cruising. Management is openly accepting slower bookings growth, lower margins, and more product investment because it knows the old model has limits. Language learning is moving from gamified exercises toward spoken, personalized, AI-assisted practice. Duolingo still has the best consumer habit machine in the category, but it has to upgrade the learning engine inside that machine.

Duolingo is still one of the strongest consumer education companies in the world, and the recent reset is strategically rational. The risk is execution timing. If DAU growth reaccelerates and speaking features improve retention, 2026 will look like a smart investment year.

If growth stays around 20% while bookings slow and AI costs rise, investors will start treating the reset as a margin-funded defense against stronger AI-native tutors.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.