Replit at $106M ARR
Jan-Erik Asplund
TL;DR: At a Sacra-estimated $106M in annual recurring revenue, Replit has been growing at a compounding monthly growth rate of 34.9% over the last 3 months, neck-and-neck with Lovable at 35.9% CMGR3 and $100M ARR. For more on Replit’s revenue expansion, 3x average revenue per customer and growth into B2B & enterprise, check out our full report and dataset.
Text-to-app has become one of the fastest growing use cases in generative AI, with Replit (last covered in June), Lovable (last covered in July), and Bolt (last covered in March) all competing to own the category. As of the end of June, Replit was growing at a Sacra-estimated 34.9% CMGR3, faster than Bolt.new (last estimated at 16.96% CMGR3 in March 2025) and neck-and-neck with Lovable at 35.86% CMGR3.
Key points from our Replit update via Sacra AI:
- Sacra estimates Replit’s annual recurring revenue (ARR) hit $106M in June 2025, up 6.6x from $16M at the end of 2024, currently raising its Series C at a $3B valuation for a 28.3x ARR multiple—compared to Cursor at $500M ARR in May 2025, valued at $9.9B (20x), Lovable at $100M ARR in July 2025, valued at $1.8B (18x), and Bolt.new at $40M ARR in February 2025, valued at $700M (18x).
- Text-to-app & vibe coding has been a massive monetization driver for Replit both in 1) driving a higher proportion of its user base to pay, going from 15K paying customers on 40 million users to 175K paying customers over the last 12 months and 2) growing ARPU from ~$192 to ~$575, expanding Replit’s market from high schoolers & people learning to code to product managers & designers at B2B & enterprise customers like Coinbase, Zillow, and Mercedes-Benz.
- Unlike Lovable (2023), Bolt.new (2024), and Cursor (2022) all launched in the past couple of years, Replit’s advantage—not unlike Vercel (founded 2015)—lies in the breadth and depth of a nearly decade-old platform (founded 2016) that spans in-browser coding, one-click deployment, and multiplayer collaboration & version control, making it more resilient to looming competition from productivity suites from Microsoft & Google and foundation labs OpenAI & Anthropic as they crowd into the space.
For more, check out this other research from our platform:
- Replit (dataset)
- Vercel (dataset)
- $172M/year Heroku of vibe coding
- Vibe coding index
- Why OpenAI wants Windsurf
- Lovable vs Bolt.new vs Cursor
- Bolt.new at $40M ARR
- Claude Code vs. Cursor
- Cursor at $200M ARR
- Cursor at $100M ARR
- Lovable (dataset)
- Bolt.new (dataset)
- Anthropic (dataset)
- OpenAI (dataset)
- Scale (dataset)
- Cursor (dataset)
