Replit, a cloud-based coding platform popular among developers and startups, has shifted its default AI model from Anthropic to OpenAI, according to CEO and co-founder Amjad Masad. Speaking at TechCrunch Disrupt, Masad explained that the decision was driven by factors including cost, performance, and the strength of the partnership with OpenAI.
Replit had previously partnered with Anthropic, relying on its Claude models to power the platform. This collaboration contributed significantly to Replit's revenue growth, with the company being one of Anthropic's top customers. However, Masad noted that OpenAI’s ChatGPT now outperforms Anthropic’s offerings on internal benchmarks related to building applications, influencing the switch.
Despite making OpenAI the default AI provider, Replit remains model agnostic. Customers can still select Anthropic’s Claude and other supported models, though the default setting likely reduces Anthropic’s usage and associated costs. Masad emphasized that the choice reflects a balance of performance, economics, and partnership quality.
Some insiders familiar with the Replit-Anthropic relationship have indicated the partnership continues, and recent company communications still reference Anthropic’s models in production. For example, a late September blog post mentioned Anthropic’s Claude Fable versions running in Replit Agent, and Replit employees have promoted Anthropic’s flagship model Opus 5.5 on social media.
Masad also shared growth insights, highlighting a rising contribution from enterprise clients. He stated that enterprise revenue grew from about 2% last year to 15% this year, with over half of Replit’s revenue now coming from B2B customers. He expects enterprise and government contracts to represent more than 50% of revenue by 2027.
On AI safety, Masad stressed the importance of a robust operational environment for AI agents, including orchestration, error recovery, security, sandboxing, networking, and identity management, to ensure safe deployment.
This update clarifies Replit’s evolving partnerships with OpenAI and Anthropic, reflecting the company’s strategic AI model choices and growth trajectory.