Wall Street spent most of 2023 and 2024 pricing Duolingo as roadkill for large language models. On September 4, 2026, Evercore ISI upgraded the language-learning company to Outperform and doubled its price target to $210, arguing that the ChatGPT-kills-Duolingo thesis has quietly broken down. Shares (NASDAQ: DUOL) traded near $154 — down 56% from their 2024 peak of $353, but up 76% from the $87.89 low set earlier this year.
The upgrade matters because Evercore is not writing about a stock that got cheap by accident. It is writing about a business that kept growing while the AI-disruption bear case was consensus.
What Evercore actually said
The new $210 target is roughly double Evercore’s prior figure and sits well above the sell-side consensus. According to Stock Analysis data compiled from 24 analysts, the average price target on DUOL is $134.29 — which implies the sell-side, on average, still sees modest downside. Evercore’s call is therefore not a piling-on move; it is an above-consensus reversal of the AI-disruption thesis that dominated buy-side conversations two years ago.
Evercore’s reasoning, per multiple summaries of the note, rests on three pillars: Duolingo’s Q2 2026 metrics showed roughly 18% year-over-year revenue growth despite the AI overhang; paid-subscriber cohorts have continued to compound; and the company’s own generative-AI features (Duolingo Max) are pushing average revenue per user higher rather than cannibalising the free-to-paid funnel.
The financial picture: growing, cash-generative, cheap-ish
Investors debating the AI-disruption thesis often lose the plot on the numbers. Here is what the actual income statement looks like.
| Fiscal year | Revenue ($M) | YoY growth | Net income ($M) |
|---|---|---|---|
| FY2022 | 369.5 | +47.3% | -59.6 |
| FY2023 | 531.1 | +43.7% | 16.1 |
| FY2024 | 748.0 | +40.8% | 88.6 |
| FY2025 | 1,038 | +38.7% | 414.1 |
| TTM (6/30/26) | 1,150 | +29.4% | 410.8 |
Revenue growth has decelerated from the mid-40s to the high-teens as the base has scaled — an expected pattern for a maturing consumer subscription business — but the profit line is doing something unusual. Net income went from a $60 million loss in 2022 to a $414 million profit in 2025, and trailing-twelve-month free cash flow now runs at roughly $408 million. The balance sheet holds $1.31 billion in cash against only $86 million of debt.
Valuation is not obviously cheap. Forward P/E sits near 58.7x and EV/EBITDA near 33.5x on Stock Analysis’s statistics page. Bulls argue that a business with 72.7% gross margins and 30%+ top-line growth deserves premium multiples; bears note that most software companies growing at that clip trade closer to a P/S of 10x, not 6.3x, which suggests the market is still discounting some AI-disruption tail risk. Evercore’s $210 target is essentially a bet that discount is too wide.
Chegg is the counterfactual
To see why the AI-resilience narrative matters, look at what happened to the other edtech name that was hit by ChatGPT at the same time.
| Metric | Duolingo (DUOL) | Chegg (CHGG) |
|---|---|---|
| Price (Sep 4, 2026) | $154.46 | $0.85 |
| Market cap | $7.23B | $94.5M |
| 52-week range | $87.89 – $353.00 | $0.45 – $1.90 |
| Revenue TTM | $1.15B | $265.5M |
| Revenue YoY | +29.4% | -47.6% |
| Net income TTM | +$410.8M | -$53M |
| Recent status | Above-consensus PT upgrade | Regained NYSE listing compliance |
Chegg’s TTM revenue has fallen roughly 48% and the stock trades below $1. Same macro backdrop, same LLM technology, same subject matter overlap with what students actually do — and one company grew profits five-fold while the other collapsed. That divergence is the substance behind Evercore’s call: AI is a real threat to some information-services businesses and a benign or even accretive shift for others.
The DUOL price arc, visualised
The stock’s journey through the AI cycle is the story in one line.
From $353 down to $87.89 is a 75% peak-to-trough drawdown. From $87.89 back to $154.46 is a 76% recovery off the low. Evercore’s $210 target would push the stock roughly 36% higher from here but still leave it 41% below the prior peak — a middle-of-the-fairway bull case rather than a return-to-euphoria call.
What could still go wrong
Three risks are worth flagging before anyone declares the AI-disruption debate over.
Growth deceleration continues. Q2 2026’s 18% year-over-year revenue growth is a step down from the 38%+ prints of the prior three fiscal years. If the next couple of quarters trend into the low double digits, the forward P/E of 58.7x becomes harder to defend. The bull case requires the deceleration to stabilise, not accelerate.
Duolingo Max monetisation. The premium AI tier is what turns generative AI from a threat into a tailwind. If attach rates disappoint on the next earnings call, the “AI is accretive to Duolingo” story starts to fray.
Competitive response from LLM providers. Nothing stops OpenAI or Anthropic from bundling a strong language-tutor mode into ChatGPT or Claude and marketing it directly to consumers. The gamified funnel and streak mechanics that Duolingo built are hard to replicate, but they are not patents.
Sources
- Yahoo Finance — DUOL quote and news feed
- Stock Analysis — Duolingo overview and consensus price target
- Stock Analysis — Duolingo annual financials
- Stock Analysis — Duolingo key statistics
- Stock Analysis — Chegg overview (comparison data)
- Duolingo Investor Relations — news releases
Disclosure: This article is for informational purposes only and is not investment advice.