EN
Back to the archive

The archive · Consumer Apps · Product decision · 2022–2023

Lensa's Magic Avatars hit No. 1, then the AI selfie wave died in 10 weeks

A four-year-old photo app bet a viral AI avatar feature could top the App Store; it did — and the category collapsed by February 2023

Prisma Labs (Lensa)

The betThat people would pay $3.99 per batch of AI selfies, turning a photo editor into a chart-topper, and that the wave would last.Live

What the business is

Lensa is a photo-editing app that builds paid AI features on top of open-source image models; Prisma Labs launched it in 2018 after its 2016 Prisma filter app went viral.

Starting capitalUndisclosed; Prisma Labs was venture-backed (Seed/Series A) after the 2016 Prisma app's viral run

How it started

Prisma Labs built Lensa in 2018 as a growing photo-editing app, but nothing moved the needle until late November 2022, when it launched 'Magic Avatars': Stable Diffusion-powered portraits generated from a user's selfies. The feature hit the top of the US App Store's Photo & Video chart within days, ahead of YouTube and Instagram ($3.99 for 50 avatars).

What happened

In one month the app amassed ~22.2M lifetime downloads and ~$29M in consumer spending, most of it earned in the viral window (Sensor Tower). The wave pulled the whole category up: by mid-December 8 of the top 100 US apps by downloads were AI-art apps, and three of the top four chart positions were AI photo editors (Apptopia). Artists and critics pushed back over training data and terms-of-service that allowed Lensa to use user images.

How it ended up

The category peaked around mid-December 2022 at 4.3M daily downloads and ~$1.8M/day in-app spending, then fell as quickly as it rose: by February 12, 2023, the same group of AI photo apps saw only ~952,000 daily downloads and ~$507,000 in spending (Apptopia). Lensa's viral moment was over within roughly ten weeks; Prisma Labs still operates the app.

Background

Lensa was Prisma Labs' second act. After the 2016 Prisma app made art filters famous, the team spent four years building Lensa into a solid but unremarkable photo editor. Nothing changed until late November 2022, when the company shipped 'Magic Avatars': upload 10–20 selfies, pay $3.99, and Stable Diffusion renders 50 AI portraits in different styles. Within days the app was No. 1 on the iOS App Store's competitive Photo & Video chart, ahead of YouTube (No. 3) and Instagram (No. 4).

The spike was enormous for a four-year-old app. Sensor Tower data cited by TechCrunch put lifetime downloads at ~22.2M and consumer spending at almost $29M since 2018, with most of that revenue concentrating in the November–December burst: 1.6M downloads in November alone, up 631% from October, and Brazil accounting for 31% of November installs (up 24,450% month over month). The wave lifted the whole category — by mid-December, Sensor Tower counted 8 AI-art apps in the top 100 US apps by downloads.

Then the trend reversed at the same speed. Apptopia data later showed the AI photo app group peaked around mid-December at 4.3M daily downloads and ~$1.8M/day in consumer spending; by February 12, 2023, the same apps saw only ~952,000 daily downloads and ~$507,000 in spending. The success was a season-long spike, not a durable business: identical apps cloned the avatar mechanic within days, and consumers who already paid had no reason to pay again.

What has to be true

  • The timing was perfect: the feature shipped right before the holidays, when people had free time, social pressure to post, and a fresh pool of selfies.
  • Open-source Stable Diffusion removed the technology moat, which was why adoption was instant — and why clones flooded the App Store within days.
  • The $3.99 price sat at the impulse-purchase threshold, converting viral attention into revenue without a subscription commitment.
  • The bet was monetization of novelty, and novelty has a half-life: one batch of avatars satisfied the need, so repeat purchase never materialized.

What can be applied

A viral feature is a loan, not equity: it converts attention into revenue only while novelty lasts. The durable product decision is what you keep after the wave retreats.

Aftermath

Prisma Labs kept operating Lensa after the wave; the app remained available on the App Store with the avatar feature priced as an in-app purchase, but it returned to the long tail of photo utilities. The industry lesson was absorbed by successors: 2023's viral AI photo runs (Remini headshots, then AI headshot apps) recycled the same playbook with the same shape, and each wave faded faster. Lensa's place in the story is as the template: the app that proved a single AI feature could top the charts in days — and that chart position buys only one quarter of attention.

Sources

spotted an error? The archive wants to know.

Your turn

You just read one. Describe what you are building, and see who is betting on the same thing.

Free account · 3 free questions · no card

Related cases