Eugenia Kuyda, the developer behind artificial intelligence companion app Replika, has secured $20 million in pre-seed funding for Wabi, a social app she describes as the “YouTube of apps.” Wabi allows anyone to create and share mini apps with basic prompts, turning creating, discovering, and hosting into one networked experience.
A social app store built from prompts and instant remixing
Wabi’s promise is simple: type what you want, and the platform whips up a working mini app from it—complete with an icon, some basic user interface elements, and any necessary data storage. Rather than publish to a traditional app store, users’ creations are posted into a social feed where friends can like, comment on, use, or remix them.
- A social app store built from prompts and instant remixing
- The investors backing Wabi’s $20 million pre-seed round
- Riding the no-code and AI wave toward social software creation
- A business model without ads and paths to monetization
- Balancing quality, safety, and scale in user-generated apps
- Why Wabi’s approach could change app discovery and sharing
The product, which is still in beta, has launched with social features including user profiles, remixing, and an Explore page that showcases popular and recent builds. Wabi will include personalization, including starter apps to fit new users’ interests, as well as feed ranking that becomes more algorithmically driven over time, says Wabi.
Early testers have shared examples of daily trivia and personal dashboards to lightweight utilities like habit trackers and simple news summaries. As with all build processes that generate code, some apps need debugging. Wabi’s team readily admits that there are limitations in the current model, but argues that ease of iteration—plus community remixing—drops enough friction to prevent the slog that often stops non-technical creators cold.
The investors backing Wabi’s $20 million pre-seed round
The $20 million round has an impressive slate of angels: AngelList co-founder Naval Ravikant, Y Combinator CEO Garry Tan, Twitch co-founder Justin Kan, Replit CEO Amjad Masad, Notion co-founder Akshay Kothari, and Neuralink co-founder DJ Seo; as well as Sarah Guo from Conviction. The backing underscores investor conviction that software creation is now moving from code to conversation—and from solo tools to social networks.
Kuyda’s track record carries weight. Worldwide, Replika has amassed tens of millions of users, showing that consumer-facing AI can maintain engagement at scale. That experience shows in Wabi’s emphasis on sticky social loops, nimble creation, and fast feedback—dynamics that drive today’s creator platforms.
Riding the no-code and AI wave toward social software creation
Wabi joins a busy market of prompt-native and vibe-coding tools such as Cursor and Lovable, as well as no-code or AI-enabled platforms like Replit, Emergent, and Bloom. It also coincides with bot and GPT storefronts, such as those in ChatGPT and Quora’s Poe. Wabi eliminates that friction point by baking in a loop: create, share, discover, and host all in one place—no packaging, no app store submission; get started and have social proof instantly.
The timing, of course, is fitting with the overall trend toward use. Gartner has predicted that over 70% of new applications will be developed via low-code or no-code approaches by 2025, propelled by a lack of developers and the ascendancy of generative AI. Wabi’s bet is that what comes next won’t be merely more accessible tools but a social fabric that transforms random users into software creators and curators.
Early momentum is apparent on social channels, where quick-build example projects have been shared by founders, designers, and investors. Product leaders at major AI labs have also chimed in, noting Wabi’s agility from idea to shareable artifact.
A business model without ads and paths to monetization
Wabi says it will not advertise, with experience showing ads nudge platforms toward dark patterns and worsen user experience. Wabi will use the new capital to hire for product and subsidize usage as it tests monetization. Likely paths include:
- A marketplace for premium mini apps
- Creator subscriptions
- Pro features like custom domains, analytics, or priority hosting
- Revenue sharing once network effects have matured
The video comparison is no accident. In the same way that early YouTube transformed from shaky video footage to professional channels, Wabi imagines a range—on one end a playful meeting of software and disposable ideas; at the other polished utilities with ongoing audiences. Unlike video, lightweight software can compound in value as creators iterate and the community remixes upgrades.
Balancing quality, safety, and scale in user-generated apps
But opening up software creation to everyone does present real challenges. Generative models can generate hallucinations; automatic scaffolding can lead to repetitive or brittle output; and community feeds can get packed with low-quality, unmaintained apps. Wabi says it is making progress on out-of-the-box reliability, better guardrails, and ranking systems that incentivize usefulness, ongoing maintenance, and originality.
Moderation, IP provenance, and the security of data will be relevant too when these apps pull from third-party sources or handle user data. Anticipate Wabi to rely on something that looks like a mix of automated checks, clear permissions, and community signals, akin to best practices from other user-generated content platforms.
Why Wabi’s approach could change app discovery and sharing
If Wabi can turn mini apps into shareable social objects—fast to build, snap to remix, easy to try—it could change how software proliferates. In lieu of trawling app stores, users could discover tools the same way they find memes and songs today: through people they trust. For Kuyda, whose earlier company helped prove out consumer AI at scale, Wabi is a bet that software itself could be the social medium.
