Base44 Unveils Proprietary AI Model to Bolster Defensibility in Vibe Coding Platform

Base44, the vibe coding platform acquired by Wix for $80 million just a year ago — when it was merely six months old with a team of eight — has begun deploying its proprietary AI model to help users build applications using natural language.

This development arrives as the AI community debates whether frontier models are optimal for every use case, and whether businesses relying on external models can maintain long-term defensibility. Base44’s latest strategy, based in the Bay Area, addresses both concerns.

Although its custom LLM is just rolling out, Base44 aims to eventually surpass frontier models. Founder Maor Shlomo notes that “training and owning the model as part of [our] entire stack allows us a lot more optimizations on latency, cost, and efficiency.”

Initially, this strategy appears designed to outpace competitors like Swedish startup Lovable, which achieved unicorn status in its Series A last summer using external LLMs. However, Shlomo anticipates that other well-scaled players will also train their own models, “at least the players that have gotten enough scale and velocity to have enough data.”

Jonathan Userovici, a general partner at VC firm Headline — which invests in AI companies like Mistral AI but not Base44 — identifies data as one of three key defensibility pillars for AI startups, alongside distribution and tech stack.

Consequently, companies with strong brands are leveraging their data and infrastructure to enhance defensibility, a pattern Base44 follows. The company states that its first LLM iteration, Base1, was trained on a dataset derived from “tens of millions of real user interactions on the platform.”

This dataset will expand alongside its rivals’. The primary competition may not be other vibe-coding startups, but rather frontier AI labs encroaching on Base44’s niche — Cursor and xAI (Grok’s parent, also owned by SpaceX) are now active in this space, and Claude Code has emerged as a significant vibe coding tool.

This gives Anthropic and other foundational AI providers access to data and feedback loops for improving app-creation models. However, Shlomo believes specialization gives Base44 an edge, predicting that “Models are progressing, but they’ll stay very general in what they can do.”

Userovici warns against underestimating frontier models, citing legal tech startup Harvey, which abandoned plans to train its own model. He does not expect applied AI companies to become frontier labs en masse but contextualizes Base44’s move within a broader trend where inference costs have become a critical factor.

Userovici notes that this cost pressure has driven changes demanded by enterprise customers. “They don’t necessarily see a [return on investment] when using the latest models for all use cases, so an entire infrastructure is being set up to do orchestration and optimization to select the right models for them so that costs don’t skyrocket while maintaining the same or similar performance across the majority of use cases.”

Although enterprise companies remain a minority among vibe coding platform users, they represent a growing share of revenue, and users of all sizes are increasingly concerned about AI costs. Base44’s decision to develop its own LLM was driven by multiple factors, with cost reduction being a key benefit.

“We want to get a model that is going to be more aligned to what we think is the right thing, is going to be more optimized to what we see users like in terms of the results we’re getting, and is going to be faster and cheaper for customers eventually than using the frontier models like Opus,” Shlomo said.

Regarding Base44 itself, cost reduction is not straightforward. In a press release, the company stated that “ownership of the model gives Base44 direct control over compute and inference spend, expected to result in a structurally stronger margin profile over time.”

Even with a delayed payoff, improved margins would benefit Base44’s parent company, which recently announced a 20% workforce reduction. In contrast, Base44 has grown its headcount since the acquisition and recently announced it had surpassed $100 million in annual recurring revenue.

This figure still trails Lovable, which reported $500 million in ARR earlier this month. However, Shlomo bets that the “huge engineering effort” to develop Base1 will solidify Base44’s position as the “only vertically integrated vibe-coding application — meaning, in Userovici’s terms, a player that owns its distribution, data, and infrastructure all at once.”