Base44 launches Base1, its own AI model for vibe coding
Base44 spent its first year building apps on other companies’ AI models. Now it has built its own. It is betting that owning the model is the only way to survive the vibe-coding gold rush.
The Tel Aviv startup lets anyone build a working app by describing it in plain language. It has now launched its first proprietary AI model, called Base1. Base44 says the model is already in production and serving users. It claims no other app-creation platform has shipped its own model rather than renting one from a frontier lab.
Base44 trained Base1 on a dataset drawn from tens of millions of real user interactions on its platform. Every app someone builds teaches the model what a good result looks like — a feedback loop rivals lack.
Base1 is a general-purpose agent built to hold conversation and write code, handling multi-turn requests, tool use, and backend operations.
Company background
Founder Maor Shlomo founded Base44 in 2024. Six months later, with a team of eight, he sold it to Wix for 150 million in annualised revenue in May 2026, two months after crossing $100 million.
Shlomo framed the launch as independence: “Having our own model means we can continue to improve it over time.” Owning the stack should eventually make inference “faster and cheaper” than frontier models like Opus, producing a “structurally stronger margin profile over time.”
Parent Wix recently cut 20% of workforce amid AI competition; Base44 has been hiring.
Competitive landscape
Rival Lovable (still on external models) claims $500M revenue. Base1 shows “competitive performance” against replaced models but does not yet beat them. Frontier labs including Anthropic (Claude Code) are moving into vibe-coding territory.
VC Jonathan Userovici (Headline) called data, distribution, and infrastructure the recipe for defensibility, but warned frontier models can undercut specialised efforts — citing Harvey abandoning its own model training plans.
The wider signal: successful AI application companies face pressure to own models, data, and costs rather than renting from frontier labs.