Replit’s Amjad Masad on the Cursor deal, fighting Apple, and why he’d rather not sell

Amjad Masad has spent the last decade building Replit, yet the past 18 months have been a transformative period. The AI coding platform has surged from $2.8 million in annual revenue in 2024 to a trajectory toward what Masad describes as a billion-dollar annual run rate.

At TechCrunch’s sold-out StrictlyVC event in San Francisco on Thursday night, we covered extensive ground, starting with the industry’s pressing question: with rival Cursor reportedly in talks for a $60 billion acquisition by SpaceX, is Replit also destined for a sale? We also discussed Replit’s net revenue retention—measuring how much existing customers expand their spending—which Masad says reaches as high as 300%. Additionally, he shared his willingness to take Apple to court over what he termed outright lies in their App Store dispute, and the possibility of Replit investing in its own customers.

On the question of independence, Masad was clear. Unlike Cursor, which he noted operates with negative 23% gross margins, he argued that Replit has the economic model to sustain its independence, even if he hasn’t completely ruled out a sale.

The following has been edited for length and clarity:

TC: Cursor’s reported SpaceX deal dominated industry conversations last week. What was your take on it?

AM: It’s challenging for an independent, smaller AI company building on foundation models, especially when burning significant cash. Some reports indicated Cursor has negative 23% margins, and if you’re also investing in model training, staying independent becomes incredibly difficult.

At Replit, we target a different customer base, allowing us to run the business more sustainably. We’ve been gross margin positive for over a year. While we are slightly more expensive, we offer significantly more value. Our audience consists largely of non-technical users who previously couldn’t create software. We provide an end-to-end platform—from the initial prompt to a deployed, scalable application. We handle security, databases, and migrations, having integrated these primitives into our platform over time.

Is Replit for sale? I assume you’re constantly in discussions with potential acquirers as part of your fiduciary duty.

Yes. We have excellent partners who occasionally raise these topics, but we aim to remain independent. I would love for Replit to stay independent. We’ve been around for 10 years, predating the acceptance of building apps from ideas alone. In 2018 at YC, we discussed creating a billion software creators, a dream that was often met with skepticism. Now, that dream is becoming reality, accelerated by our agentic coding experience launched in September 2024. It feels like we can take this much further.

You collaborate closely with Anthropic, Google, and OpenAI. If you had to rank them, who is performing best?

Anthropic remains undefeated in the core agentic loop, offering the best tool calling and maintaining agent coherence longer. GPT-5 is catching up rapidly. Google’s Flash models are exceptional in price-performance, currently outperforming open-source options for speed and cost. We use all three, and I wouldn’t discount newer labs either. Reflection AI is releasing open-source models that are receiving positive feedback, and Chinese models like Kimi are impressive, matching Anthropic’s January capabilities and staying only about three months behind.

In enterprise deal competitions, what gives you the winning edge?

Most of our sales are inbound or organic, driven by product-led growth. We’ve acquired customers like Zillow and Meta through organic adoption leading to enterprise plan purchases. When deals do go top-down with formal evaluations, we usually win on product quality. Even if we lack a specific feature, Replit wins in security discussions with C-suite and IT teams. Many vibe-coding tools generate websites connected to external databases, which complicates security by exposing databases publicly and requiring complex row-level security configurations—especially difficult for non-technical users. Replit’s full-stack approach, with databases built into the project and not publicly exposed, makes applications inherently more secure.

We’ve also spent 10 years combating crypto scammers and hackers, making our cybersecurity function comparable to a dedicated security startup. Every app deployed on Replit creates a new, isolated project on Google Cloud, inheriting Google’s robust security model.

Can we discuss churn? How do you retain customers if the best prototypes are eventually rebuilt into a company’s existing stack?

Churn is very low, and net retention is exceptionally high—reaching 300% in some cases. Customers often report that when engineers attempt to rebuild an app into their own stack, the result is frequently worse. Once enterprises become comfortable with the full Replit stack, particularly when we set up single-tenant environments, they keep the apps on Replit. For example, Bain & Company replaced Tableau and Power BI with Replit and Databricks.

There’s growing concern about AI bloat—non-technical users generate more code and consume more tokens. This benefits you due to usage-based fees, but what about your customers?

We don’t see much regrettable spend. Enterprises are highly ROI-focused and report significant returns. Most feel the investment is entirely justified, often yielding returns one to three orders of magnitude higher. If a company spends $100,000 monthly with Replit, they typically generate $2 million to $10 million in return.

Let’s talk about Apple. Another rival, Lovable, recently got an app-building tool approved by the App Store. Replit has been in App Store purgatory, with Apple blocking your updates for months. How much does this impact you?

It’s not critical—we could lose the app without significantly affecting our business. However, it’s an app people genuinely love. We’ve been on the App Store for four years. Students in underprivileged communities learn to code on Replit using Android devices, and executives use it in meetings.

We believe Replit was blocked while others weren’t because we enable iOS app development. When we launched this capability in December, charts circulated showing how many apps were entering the App Store through us. We suspect Apple feels threatened by this.

Apple’s stated reason is that you’re downloading new code to the device after approval, violating their guidelines.

That’s a lie. We can prove it in court if necessary.

Will that happen?

I hope not. I’m an Apple fan and would love to collaborate and build something great together. We’re happy to direct customers to Xcode, Apple’s own development environment. But you can’t run a marketplace accessible to a billion people while making discriminatory or whimsical decisions.

Are you considering investing in your own customers for equity, similar to Nvidia, OpenAI, and others?

We’ve considered this seriously. I’ve personally invested in several startups that started on Replit before generating revenue. For instance, Magic School began when a teacher used his time during COVID to learn vibe coding and built an AI app for other teachers. He identified a problem: teacher burnout in America. Using AI to reduce workload, he achieved $20 million in revenue in his first year. Other companies starting on Replit are now valued at half a billion dollars. The entrepreneurship on Replit is genuinely exciting. We integrated with Stripe a few months ago, and transactions flowing through Replit are growing triple digits month over month. Soon, our customers will be generating more revenue than we do.