Replit announced a $400 million Series D round on Wednesday, pushing its valuation to $9 billion. Six months ago the company was worth $3 billion. Georgian Partners led the round, with G Squared, Prysm Capital, Coatue, Andreessen Horowitz, Craft Ventures, Y Combinator, and Qatar Investment Authority piling in. Shaquille O'Neal and Jared Leto also invested, because apparently that's how funding rounds work now.
The math problem
Replit won't say what its current annualized recurring revenue is. The company told Forbes it hopes to hit $1 billion ARR by year-end, which is a big number to throw around without showing your work. Back in September, during its $250 million raise, Replit reported $150 million in annualized revenue. Getting from $150 million to $1 billion in roughly a year means something like 6x growth. Possible in the vibe-coding gold rush, but aspirational doesn't mean inevitable.
The new funding makes CEO Amjad Masad a billionaire for the first time, worth an estimated $2 billion according to Forbes. He once turned down a $1 billion acquisition offer for the whole company. That decision looks vindicated, at least on paper.
What Agent 4 actually does
Alongside the funding news, Replit shipped Agent 4, the latest version of its AI coding agent. The headline feature is parallel execution: multiple AI agents can work on different parts of a project simultaneously, then merge their changes. Think authentication on one thread, front-end design on another, database setup on a third.
There's also an infinite design canvas that lets users generate multiple UI variants at once, pick a favorite, and push it directly into the app. Masad told Inc. that Agent 4 can "vibe code a startup from scratch," which is the kind of claim that sounds great in a press cycle and harder to evaluate in practice. He described Replit as "the cockpit or the launch control of your business." The parallel agents feature is limited to Pro and Enterprise users, though Core users get temporary access as a launch promotion.
So who's actually winning?
Here's the uncomfortable context Replit's announcement doesn't dwell on. Claude Code, Anthropic's coding product, reportedly hit $2.5 billion in annualized revenue, according to the Forbes piece. Cursor's parent company Anysphere is at $2 billion ARR. Even Lovable, with just 146 employees, claims $400 million ARR. Replit has 50 million users and a presence in 85% of Fortune 500 companies ("users from" those companies, which could mean one intern), but the revenue gap with Claude Code is enormous.
Replit's bet is that it's building for a different audience. Professional developers already have Cursor, Claude Code, GitHub Copilot with its 20-million-plus user base. Masad has spent nine years pivoting Replit away from that crowd and toward non-technical users: marketers, sales staff, small business owners. People who describe what they want in plain English and let the AI figure out the rest.
That positioning makes strategic sense. It also means Replit is competing less on raw coding power and more on interface simplicity, which is a race where incumbents can catch up fast. Anthropic's Claude Cowork already targets enterprise teams with a more accessible interface than Claude Code.
Where the money goes
The capital will fund international expansion, with a focus on Asia and the Middle East, plus growing Replit's go-to-market team. Margaret Wu, the lead investor at Georgian Partners, framed the opportunity around the expansion of who gets to build software. The practical question is whether Replit can convert its 50 million users (many of them students and hobbyists) into paying enterprise accounts fast enough to justify a $9 billion price tag.
Enterprise customers already on the platform include Zillow, LabCorp, PayPal, Atlassian, and Adobe, according to the company. UKG, a human capital management platform, said using Replit cut prototype development from weeks to days. Those are real use cases. Whether they add up to $1 billion in annual revenue by December is a separate question entirely.
Replit was founded in 2016, and it took nine years of grinding before the vibe-coding wave carried it to relevance. As Masad previously told TechCrunch, the controversial decision to abandon professional developers is what finally unlocked growth. The next test is whether that same audience can sustain a $9 billion company.
