Wabi

Wabi raised fast and staffed its small team mostly with founders. It raised a $20M pre-seed led by Andreessen Horowitz one month after its October 2025 iOS public beta. Nine of its 14 people are founders.

Kuyda, a second-time founder after Replika, says a small team with many AI agents is bounded only by imagination, and the host notes founders suit a product needing speed. a16z's Anish Acharya cites Wabi as an example of consumer AI built on coding agents.

Subversive

Ordinary apps must be businesses, which pushes them toward dark patterns. Eugenia Kuyda argues every app has to justify itself as a business, so apps that would not make money never get built and the rest push users to click, buy or scroll, with tricks like a tiny light-gray paywall close button.

Wabi moves auth, security, back end and payments to the platform layer so no single mini app has to be a business. In her words, the question apps end up asking is "how do we trick people to do things that's good for our business but not necessarily good for them."

Lenny's Podcast▶ 1:05:25

AI made building apps nearly free, but nowhere existed to use them. A simple weight tracker takes about a minute to make, yet publishing it still means a developer account and an app store. Only about 10 million iOS and Android developers serve billions of people.

Wabi's answer is one place to discover, remix and use small personal apps. Kuyda's stated goal is software that is "ephemeral, malleable, personalized, alive, ever evolving."

Lenny's Podcast

Value comes from many mini apps sharing context, not one app. Wabi apps share context and memory and can talk to each other. Adding a New York trip to a kids calendar leads a kids-events app to suggest events in New York that week.

Kuyda's own fitness coach reads across her weightlifting tracker and weight tracker to suggest additions. "The power is in the constellation, not just one mini app."

Lenny's Podcast▶ 54:55

Agentic apps that reach out make the platform feel alive. Wabi is blurring agents and apps so apps can message the user and propose changes, such as new kids' events they found.

Showing what friends are using adds the social layer the App Store lacks. Kuyda plays a game with her mom on Wabi because the platform gives them a leaderboard. "A consumer social product should feel alive."

Lenny's Podcast▶ 35:15

Define activation as one useful app someone returns to, not creation. Wabi's main goal is getting a new user to one first useful app. One or two apps used daily or weekly make a daily active user, so nobody needs to browse 50 apps on the explore page.

Many people have no idea what to build, so the job is to surface an app they will use and come back to.

Lenny's Podcast▶ 12:00

Lead new users to remix existing apps instead of starting blank. Kuyda expects about 90% of people to create nothing in their first days, so Wabi surfaces apps to try and then nudges users to edit or remix one.

Remixing removes the blank-page problem because someone else already designed the functionality. She turned a painting-collection tracker into a Pokemon card tracker with one sentence of instruction. Wabi gives no retention data showing remixing beats building from scratch.

Lenny's Podcast▶ 15:05

Seed the horizontal platform with the categories people actually pick. Early Wabi activity clusters around trackers of all types (supplements, food, reading), then games and quizzes, then learning and entertainment, plus family tools like vacation votes and packing lists.

Highlighting these to new users avoids the blank page. Most users are around 30, though demographics spread widely. Kuyda's own apps are plain daily tools: lifting and weight trackers, a personal CRM, a daily poem.

Lenny's Podcast▶ 32:30

Wabi watches retention and K factor because sharing is its growth engine. The metrics Kuyda tracks most closely are "just retention and the K factor." People use apps with others, share them, and creators can bring their fan base.

This is a reaction to Replika, a single-player product where talking to an AI was stigmatized, so word of mouth was hard and growth came from press: from a couple million users in early 2018 to more than 40 million.

Lenny's Podcast▶ 1:06:30

Sell agency to a broad audience, not prompt-to-app technology. Outside the tech bubble most people don't know they can create or discover apps, so Wabi frames itself as setting software free and turning consumers into creators. Its short Instagram videos consistently get half a million to a million views.

Kuyda carries this from Replika, where a true human story spread when users wouldn't: "the most viral thing you can create is a story that people want to tell each other," and it can't be fake.

Lenny's Podcast▶ 27:30

Plan to charge users the running cost and pay creators a cut. Since many Wabi apps run AI inference, Kuyda wants users to see what each mini app costs and pay for it, "a little bit of pocket money to run" a small army of apps. Creators whose apps are remixed or heavily used would get a share.

This is still a hypothesis with no revenue figures, and ads are not fully ruled out.

Lenny's Podcast▶ 1:08:50

Wabi raised fast and staffed its small team mostly with founders. It raised a $20M pre-seed led by Andreessen Horowitz one month after its October 2025 iOS public beta. Nine of its 14 people are founders.

Kuyda, a second-time founder after Replika, says a small team with many AI agents is bounded only by imagination, and the host notes founders suit a product needing speed. a16z's Anish Acharya cites Wabi as an example of consumer AI built on coding agents.

Lenny's Podcast▶ 1:16:10