Generator ownership determines winners
Lovable at $84M ARR growing 36% MoM
This market is being shaped less by who can generate code, and more by what each company already owns underneath the generator. Replit starts with a browser IDE and collaboration layer, so its wedge is teams already writing and editing code together. Vercel starts with deployment and hosting, so it can turn a generated app into ongoing infrastructure spend. Bolt.new and Lovable started AI first, which made them faster to ship a simple prompt to app experience and faster to grow with non developers.
-
Replit added Agent on top of an existing multiplayer coding product, after reaching 20M users by 2023 and building collaboration, databases, version control, and one click deployment into the core workflow. That means its natural move is from hobbyist coding into team software creation and then into internal tools and enterprise use cases.
-
Vercel is different because the app builder is attached to a PaaS business. A user can generate a Next.js app in v0, deploy it straight to Vercel, and keep paying for compute, bandwidth, storage, seats, and enterprise hosting. The generator is customer acquisition for the infrastructure business.
-
Bolt.new and Lovable helped define the AI native app builder category by making prompt to working app the main experience from day one. Bolt added mobile support and Figma import, while Lovable grew out of GPT Engineer and fits a two step workflow where users prototype in the browser, then move the code into tools like Cursor for deeper editing.
From here, the category will split by retention engine. Platform based players like Vercel and Replit are best positioned to hold onto apps after creation through hosting, collaboration, and team workflows. AI native upstarts like Lovable and Bolt will keep pushing on speed, simplicity, and distribution, then add multiplayer, ecosystems, and more of the stack to avoid becoming just the first draft tool.