How two of the top next-generation developer platforms spent years building the foundation for their AI breakout moments—and what that tells us about “sudden” success in the age of AI
The headlines make it sound simple: Vercel raised $300M Series F at a $9.3B valuation in September 2025 (co-led by Accel and GIC, with participation from BlackRock, StepStone, and Khosla Ventures), with revenue doubling and users skyrocketing after launching v0, their AI-powered development agent. Replit went from $2.8M to $150M in annualized revenue in less than a year after releasing Replit Agent, securing a $250M Series D at $3B valuation (led by Prysm Capital, with Amex Ventures and Google’s AI Futures Fund joining as strategic investors).
These are the kind of hockey-stick growth stories that make VCs salivate and founders feel inadequate. But here’s what the breathless press releases don’t tell you: both companies spent nearly a decade preparing for this “overnight” success.
The Real Story: Ten Years in the Making
Vercel’s Journey: 2015-2025
Guillermo Rauch founded Vercel (originally ZEIT) in 2015 with a vision to make web deployment as simple as typing “now” in the command line. But the path wasn’t smooth:
- 2015-2020: The company struggled with investor perception. As one early investor candidly admitted, they were “glad to resume business with Guillermo, after he returned to the path of entrepreneurship”—suggesting the company wasn’t initially seen as a serious startup
- Revenue growth finally began in 2019, but not “AI crazy” growth yet: $1M in 2019, $5M in 2020, $21M in 2021
- October 2016: Released Next.js as open source—the framework that would become their trojan horse
- The platform play: Spent years building deployment infrastructure, developer experience, and framework optimization while Next.js slowly gained traction
- April 2020: Rebranded from ZEIT to Vercel, a recognition that the initial brand wasn’t working
- From $2.5B (2023) → $3.25B (May 2024) → $9.3B (Sept 2025). That’s a 45x revenue multiple on $200M ARR—stratospheric even for AI compan
Fast forward to 2025: Vercel hit approximately $200M ARR by mid-year. Their AI SDK gets 3 million downloads per week. v0 has reached 3.5 million users, with Teams and Enterprise accounts representing over 50% of v0 revenue. The company doubled its user base and increased revenue 82% year-over-year.
The result: September 2025 Series F — $300M at $9.3B valuation (co-led by Accel and GIC), nearly tripling from their $3.25B valuation just 15 months earlier.
Replit’s Journey: 2011-2025
Amjad Masad’s story is even more striking. He started building repl.it as an open-source side project in 2011 while working at Yahoo in Jordan, frustrated by the pain of setting up development environments in internet cafes.
- 2011-2016: Built the technology as a passion project, with 10K users organically appearing despite minimal attention
- 2016: Officially founded Replit as a company after leaving Facebook
- 2016-2024: Eight years of grinding. The company reached $2.8M ARR by 2021 and then… stayed there for 3-4 years
- The brutal plateau: “We had reached that $2.83 million back in ’21, maybe,” Masad recalled. “And so this is how painful it’s been. We’ve been hovering around the same revenue for like four or five years”
- Rejected by YC four times before Paul Graham found them on Hacker News and had Sam Altman reach out
- Mid-2024: Had to cut half the staff—down to 65 people—to refocus on AI
- September 2024: Launched Replit Agent
- From $1.16B (2023) → $3B (Sept 2025). Revenue grew 50x from $2.8M to $150M ARR during that period.
What happened next? Revenue exploded to $70M ARR by April 2025, then $100M by June 2025, and $150M by September 2025. The company grew from 22.5 million developers to 40 million users. They now have $350M in the bank despite barely touching their previous funding.
The result: September 2025 Series D — $250M at $3B valuation (led by Prysm Capital), nearly tripling from their $1.16B valuation in 2023.
The Pattern: Infrastructure First, AI Second
Here’s what makes these stories instructive for SaaS founders: neither company built their AI products in a vacuum. They spent years building the infrastructure that made their AI products possible.
Vercel’s Foundation:
- A decade of perfecting Next.js adoption (now used by millions of developers weekly)
- Deep relationships with enterprises like OpenAI, Anthropic, PayPal, Nike, and Walmart
- Infrastructure for instant deployment and preview environments
- A developer-first brand built through years of community engagement
Replit’s Foundation:
- 13 years of building browser-based development environments
- Expertise in running multiple programming languages in the cloud
- One-click deployment infrastructure
- A loyal community of 22+ million developers who understood the product
When the breakthrough AI models arrived in 2024 (particularly Claude 3.5 Sonnet and GPT-4), both companies were positioned to move fast because they already had:
- Distribution: Millions of existing users who could immediately try the new AI features
- Infrastructure: The technical capability to actually deploy and run AI-generated applications
- Domain expertise: Deep understanding of what developers need beyond just code generation
- Trust: Years of reliability that made enterprises comfortable adopting AI-powered tools
The Timing Factor: Why Now?
Both CEOs will tell you timing was everything—but it was preparedness meeting opportunity.
Masad was explicit about this: “We started Replit officially in 2016 because this thing just wouldn’t die and kind of exploded itself into the world.” He had been thinking about democratizing programming since he was learning to code in internet cafes in Jordan as a teenager.
Rauch had been obsessed with developer experience since creating Socket.IO and building JavaScript infrastructure at his previous companies. The vision for Vercel existed years before the technology to fully realize it.
The AI breakthrough didn’t create their success—it unlocked the value of infrastructure they’d been building for a decade.
The Competitive Moat Question
Both companies face existential questions about moats:
For Vercel:
- Competitors like Bolt.new ($40M ARR), Lovable ($70M ARR by June 2025), Cursor ($200M ARR), and traditional cloud providers are all attacking different parts of their value chain
- Their moat is the integration: v0 generates Next.js code that deploys instantly on Vercel infrastructure with built-in security, monitoring, and scaling
- They’re becoming the “omakase” option—everything you need in one opinionated package
For Replit:
- The foundational model companies (Anthropic, OpenAI) are launching their own coding tools
- Major competitors like Cursor raised massive rounds and are growing fast
- Replit’s bet: targeting non-technical users and building sophisticated deployment infrastructure that model companies don’t prioritize
- As Masad put it: “We don’t care about professional coders anymore”—they’re going after the billion people who want to build software but can’t code
What B2B Founders Can Learn
1. The “overnight success” narrative can be dangerous
Both companies spent the better part of a decade building before their breakout year. Vercel grew from $1M to $200M over six years before AI accelerated growth. Replit stayed at $2.8M for 3-4 years before exploding to $150M.
If you’re building infrastructure or developer tools, patience isn’t optional—it’s the price of admission.
2. Open source as distribution
Next.js and repl.it (the open source project) were both free, open-source projects that built massive communities before commercialization. Vercel’s Next.js is downloaded 200 million times per week. Replit had millions of users before they figured out monetization.
The open source community became their sales force when the AI moment arrived.
3. Capital efficiency matters—until it doesn’t
Replit raised $100M in 2023 and hadn’t touched it by their 2025 raise. Masad jokes he needs to learn to be “less frugal” after growing up watching his refugee father struggle. But that frugality meant they had runway when they needed to cut staff and refocus.
Once product-market fit hit, both companies raised massive rounds at eye-watering valuations. The market pays premiums for proven execution.
4. Platform plays take time
Both companies are platforms, not point solutions. Vercel isn’t just a deployment tool—it’s infrastructure, framework, AI SDK, and developer experience. Replit isn’t just an IDE—it’s coding environment, deployment, database, and AI agent.
Building platforms requires more upfront investment and takes longer to see returns, but the moats are deeper when you succeed.
5. The AI moment requires existing distribution
Neither company could have achieved this growth starting from zero in 2024. The AI tools are table stakes—everyone has access to Claude, GPT-4, and other models. The difference is:
- Vercel already had millions of developers deploying to their platform
- Replit already had tens of millions of users in their IDE
- Both had enterprise relationships they could upsell
Your AI feature is only as good as your ability to get it in front of users.
The Path Forward: Sustaining Hypergrowth
Both companies now face the classic scale-up challenges:
Leadership builds:
- Vercel brought in executives from Stripe (COO), Redis (CMO), Capital One (SVP Product), HashiCorp (CAO), and IBM (CTO Security)
- Replit is scaling from 110 employees with Masad admitting they need to learn to spend money
Product expansion:
- Vercel launched v0 Mobile (enabling voice and camera-based development), AI Gateway, and AI Sandbox
- Replit launched Agent 3 (their most autonomous agent), Agent v2, and Mobile Agent
Market positioning:
- Vercel is going upmarket to enterprise with security and compliance features
- Replit is staying true to democratization but adding enterprise-grade tools
The Takeaway: Infrastructure Is Destiny
The most important lesson from Vercel and Replit isn’t about AI—it’s about infrastructure.
Rauch spent a decade making deployment simple. Masad spent 13 years making coding accessible. When the AI models got good enough, their infrastructure was ready to support a new paradigm of software creation.
The “overnight” success happened because they built the rails before the train arrived.
For SaaS founders watching the AI wave and wondering how to ride it: you probably can’t build the Vercel or Replit of your category in the next six months. But you can start building the infrastructure today that will let you capitalize when the next breakthrough moment arrives in your space.
The companies that win the AI era won’t be the ones that slap ChatGPT onto their existing product. They’ll be the ones who spent years building the foundation that makes AI-powered experiences possible.
Vercel and Replit are “overnight” successes the same way that Amazon was an “overnight” success after spending years building fulfillment centers, or Netflix was an “overnight” success after spending years building streaming infrastructure.
The timing was perfect. But only because they’d been preparing for a decade.
Not All Success Stories Are Remotely Overnight
Vercel: 10 years from founding to $9.3B valuation. $200M ARR growing 80%+ annually. Built on the back of Next.js adoption and years of developer experience optimization.
Replit: 13 years from first code to $3B valuation. $150M ARR after hovering at $2.8M for years. Built on the back of browser-based development environments and a massive user community.
Both companies are now racing to build the future of software development. But they’re only in this race because they spent years building the track first.
That’s not an overnight success story. That’s a masterclass in patience, persistence, and preparation meeting opportunity.
