Replit, an AI coding platform popular among startup founders and developers, is no longer working with Anthropic, according to CEO and co-founder Amjad Masad. Speaking onstage at TechCrunch Disrupt, Masad stated that the decision was driven by factors including cost, partnership dynamics, and performance.
Shift to OpenAI as Default
While Replit announced a partnership with OpenAI in August, the status of its relationship with Anthropic had remained unclear until Masad’s remarks. Masad noted that ChatGPT is currently performing better, pointing to internal and public benchmarks where models like Sol and Astra excel at building applications from scratch.
Following the event, Masad clarified that Anthropic models have not been entirely removed from the platform. Customers can still choose to access Anthropic’s Claude, but OpenAI’s ChatGPT now serves as the default. Because most platform users stick with the default option rather than selecting a model manually, the shift could significantly reduce Replit’s spending with Anthropic. Previously, Replit relied on Anthropic’s Claude models, a partnership that helped drive significant revenue growth and made Replit a top-five customer for Anthropic.
Conflicting Accounts and Ongoing Promotion
Despite the CEO’s on-stage statements, insiders familiar with the partnership disputed the claim that the relationship had ended, maintaining that the partnership remained in place. Public evidence on Replit’s website, blog posts, and social media also showed the company promoting Anthropic products shortly before the announcement, including references to Claude Fable models in late September.
Clarifying the company’s current stance, Masad reiterated that Replit remains model agnostic. He stated that OpenAI is now the preferred partner for powering Replit Agent due to performance, economics, and the strength of the partnership.