Replit CEO says OpenAI is now the coding platform's default model, though Anthropic's Claude models remain available

What happened

Replit, an AI coding platform popular with startup founders and developers, has made OpenAI's ChatGPT the default model on its service instead of Anthropic's models, TechCrunch reported. CEO and co-founder Amjad Masad first described the change onstage at TechCrunch Disrupt, saying Replit was no longer working with Anthropic.

Masad later clarified that customers can still pick Anthropic models on the platform. OpenAI is now the default, according to TechCrunch. Replit sells a cloud-based platform where users build and deploy apps directly in a web browser.

Why Replit says it switched

Cost, performance and partnership

Masad said several factors drove the decision, including cost. "I think ChatGPT is better now," he said onstage. He said Replit uses an internal benchmark to measure things like performance and also publishes a public benchmark.

He said models called Sol and Astra have consistently done better at building what he described as vibe-coded applications from scratch. Replit announced a partnership with OpenAI in August. TechCrunch noted that nothing at that time suggested the company was ending its work with Anthropic.

Anthropic models remain available

"Replit remains model agnostic," Masad told TechCrunch over the weekend. He said users can choose Anthropic's Claude or other supported models, and that OpenAI is now Replit's preferred partner for powering Replit Agent. TechCrunch reported that the change could still significantly reduce Replit's spending with Anthropic, because most customers stay with the default model rather than choosing one themselves.

Questions about the Anthropic relationship

Insiders dispute the account

Insiders familiar with the partnership rejected Masad's onstage comments, according to TechCrunch. One of them, who asked not to be named, said the partnership between the two companies was still in place.

Before the OpenAI deal, Replit ran its platform on Anthropic's Claude models. That partnership was said to have helped Replit grow its revenue significantly. Masad said onstage that Replit had been one of Anthropic's top five customers until recently.

Recent promotion of Claude models

TechCrunch found evidence on Replit's website, blog and social media that the company was promoting Anthropic products as recently as a few weeks ago. A Replit blog post in late September listed Anthropic's Claude Fable 5 and Claude Fable 5.1 as models running in Replit Agent's production environment. Around the same time, a Replit product employee posted on X about using Anthropic's flagship model, Opus 5.5, and promoting its capabilities.

Replit's business shift

Masad also said enterprise customers now make up a larger share of Replit's revenue. He said enterprise was about 2% of revenue last year and is about 15% this year. He added that more than 50% of revenue comes from business-to-business sales. He said the company is not yet sharing annual recurring revenue figures.

Backstage, Masad told TechCrunch he expects more than half of Replit's revenue to come from enterprise and government customers in 2027. He also spoke about AI safety. He said AI agents need a "habitat" with error recovery, sandboxing, security, and identity and authorization controls before they can be deployed safely. For more coverage of the companies building AI tools, see our AI and tech section.

What to watch next

• Whether Replit or Anthropic clarifies the status of their partnership, which Masad and a person familiar with the deal have described differently.
• How much Replit's spending with Anthropic falls now that OpenAI is the default model for most users.
• Whether Replit reaches Masad's expectation that more than half of its revenue will come from enterprise and government customers in 2027.
• Whether Replit releases annual recurring revenue figures, which Masad said the company is not sharing yet.