Replit, the popular AI-powered coding platform that has captured the loyalty of startup founders, software engineers, and hobbyist developers alike, has severed its working relationship with AI lab Anthropic. The surprising revelation was delivered by Amjad Masad, the CEO and co-founder of Replit, while speaking onstage at the TechCrunch Disrupt conference.
According to Masad, the decision to part ways with Anthropic was driven by a confluence of factors, chief among them being cost considerations, partnership dynamics, and overall performance metrics. Although Replit formally announced a strategic partnership with OpenAI back in August, the announcement at the time did not explicitly signal the end of its foundational ties with Anthropic, leaving many in the tech industry under the impression that multi-model support or a broader ecosystem approach was still intact.
However, Masad’s statements at TechCrunch Disrupt painted a much clearer picture of a shifting alliance. "I think ChatGPT is better now," Masad told the audience, explaining that Replit relies on a rigorous internal benchmarking system to continuously evaluate the performance, speed, and reliability of various underlying large language models. "We publish a public benchmark as well, and consistently, Sol and Astra perform better on building vibe-coded applications from scratch. It was a decision based on both cost, partnership, and performance. And for all these reasons, we just felt like they were a better partner for now."
For context, Replit operates a comprehensive, cloud-based software development platform that allows users to write, build, test, and deploy software applications directly from within their web browsers. Under its previous partnership with Anthropic, Replit leveraged the AI lab’s sophisticated Claude models to power core functionalities of its platform. Industry observers and financial analysts have widely credited that collaboration with helping Replit achieve exponential business growth, with reports indicating that the platform managed to scale its revenue tenfold following the integration of Anthropic’s technology.
Yet, Masad’s public declaration quickly generated friction and conflicting accounts from industry insiders close to the situation. Following his remarks, sources familiar with the partnership between Anthropic and Replit pushed back against the narrative of a total and immediate separation. At least one knowledgeable source, speaking on the condition of anonymity, insisted that the partnership was still very much in place, complicating the timeline and exact nature of the platform’s current vendor relationships.
Further complicating the picture is verifiable documentation showing that Replit was actively integrating and utilizing Anthropic products in its production environments as recently as September. For instance, a technical blog post published on Replit’s official website in September explicitly cites Anthropic’s Claude Fable 5 and Claude Fable 5.1 as active models running within the production environment of the Replit Agent. In light of these conflicting accounts, TechCrunch reached back out to Replit leadership for further clarification regarding the exact status of the Anthropic integration, though an official response had not yet been provided at the time of publication.
Beyond the shifting landscape of its underlying AI model providers, Masad used his appearance at TechCrunch Disrupt to shed light on Replit’s broader financial health and commercial evolution. While declining to disclose exact Annual Recurring Revenue (ARR) figures, the CEO highlighted a dramatic shift in the company’s customer acquisition and monetization models, pointing out that enterprise customers are rapidly becoming a major driver of top-line growth.
"We’re not sharing ARR numbers just yet, but in terms of growth, like, I would say, enterprise was like 2% of revenue last year," Masad shared with the Disrupt audience. Highlighting the speed of the company’s B2B transformation, he added, "This year, I would say like 15% of our revenue is enterprise contracts, and more than 50% of our revenue is B2B."
The discussion also touched upon the broader technical and operational requirements needed as software development increasingly shifts toward autonomous AI agents. Masad emphasized that as artificial intelligence takes on more complex coding and engineering tasks, the surrounding digital infrastructure must evolve to ensure safety, reliability, and security. Drawing an analogy to human workplaces, he stressed that advanced AI tools require structured, secure environments to function effectively.
"In the same way that us humans need a certain environment to work — you need a nice office, maybe nice good food in the office, air-conditioning, computer — agents need a habitat for them to exist and do good work," Masad said. "So that habitat has good orchestration, good error recovery, good safety and security and sandboxing, good networking, good identity and authorization. And once you figure all of that out, then you can deploy AI safely."