Replit Ends Partnership with Anthropic, Focuses on OpenAI Models

Image: TechCrunch · Source

Replit’s CEO Amjad Masad announced the company is no longer working with Anthropic due to cost, performance, and partnership considerations, shifting focus to OpenAI’s models for its AI coding platform.

Replit, a popular AI-powered coding platform favored by developers and startup founders, has officially ended its partnership with Anthropic, according to CEO and co-founder Amjad Masad. The announcement was made during TechCrunch Disrupt on October 16, 2026. Masad cited multiple factors influencing the decision, including the cost of collaboration and performance metrics.

Previously, Replit used Anthropic’s Claude AI models to power its platform, which allowed users to develop and deploy applications directly from their browsers. The partnership with Anthropic had contributed to substantial revenue growth for Replit. Despite this, Masad explained that Replit now finds OpenAI’s ChatGPT and related models to offer better performance, especially in building vibe-coded applications from scratch, based on internal and public benchmarks.

Masad stated, “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.” Although a September blog post indicated ongoing use of Anthropic’s Claude Fable models in production environments, Masad’s comments mark a clear shift in Replit’s AI model strategy.

Some insiders familiar with the previous partnership expressed skepticism about the termination, suggesting it might still be active. However, Replit’s current direction leans heavily on OpenAI technologies following a partnership established in August 2026.

Beyond AI partnerships, Masad shared insights into Replit’s growth trajectory, highlighting a significant increase in enterprise contracts—from 2% of revenue last year to 15% currently. Overall, more than half of Replit’s revenue now stems from B2B customers.

Masad also emphasized the importance of creating a safe and well-orchestrated environment for deploying AI, comparing it to the necessity of a conducive workspace for humans. He mentioned factors like error recovery, security, sandboxing, and proper authorization as critical for safely operating AI agents.

Replit’s pivot away from Anthropic toward OpenAI models reflects the fast-evolving AI development landscape and the company’s focus on optimizing performance and cost-efficiency for its growing user base.