Lifestyle
Replit CEO says it no longer works with Anthropic, but a source says the partnership continues
Replit CEO Amjad Masad said at TechCrunch Disrupt that the AI coding platform no longer works with Claude maker Anthropic, citing cost and performance. TechCrunch cites a source who says the partnership is still active. Replit has not yet clarified, and the impact on users is unclear.
About 5 min read
What happened
TechCrunch reports that Replit CEO and co-founder Amjad Masad said on stage at TechCrunch Disrupt on Thursday that Replit no longer works with Anthropic, the company behind the Claude family of AI models. According to TechCrunch, Masad said the decision had several reasons, including cost.
TechCrunch quotes Masad as saying, "I think ChatGPT is better now," and "It was a decision based on both cost, partnership, and performance." TechCrunch adds that Masad said the company uses an internal benchmark to evaluate performance and other metrics. A benchmark is a standardized test used to compare how different AI models perform. He said the company also publishes a public benchmark, and that Sol and Astra consistently do better at building vibe-coded apps from scratch. Astra refers to GPT-6 Astra, an OpenAI model that Replit uses. Vibe-coded apps are apps whose code is largely generated by AI.
TechCrunch notes that Replit announced a partnership with OpenAI in August, but there was previously no sign that it had stopped working with Anthropic. The claim that the partnership has ended currently comes mainly from TechCrunch's report of Masad's remarks.
Accounts are not consistent
TechCrunch reports that a source who asked not to be named said the partnership is still active. TechCrunch also notes that a September blog post listed Anthropic's Claude Fable 5 and Claude Fable 5.1 among the models running in production for Replit Agent. Replit Agent is the AI assistant on the Replit platform that builds software for users. Production means the live service actually offered to users.
In its own engineering blog post of September 29, Replit said it sampled each model in Agent production for one week: Fable 5 in August 2026, and Fable 5.1 and GPT-6 Astra in September. This only describes usage at that time and does not indicate the current status. TechCrunch says it has asked Replit for clarification again and will update its report if it receives a reply.
| Item | Current claim | Source |
|---|---|---|
| Partnership status | Masad says Replit no longer works with Anthropic | TechCrunch report of Masad's remarks |
| Differing view | An anonymous source says the partnership is still active | TechCrunch report |
| Recent usage record | A September post lists Claude Fable 5 and Fable 5.1 running in Agent production | TechCrunch; Replit blog |
| OpenAI relationship | Replit announced a partnership with OpenAI in August | TechCrunch; Replit blog |
Replit's OpenAI moves and business background
In an August 18, 2026 blog post, Replit announced Free Mode, saying it is powered by OpenAI's GPT-5.6 Luna and that Core subscribers can use it for $20 per month. This is Replit's own statement.
In its September 29 post, Replit published DeepSWE v1.1 results. DeepSWE is a software engineering test of whether AI can make long-running changes to open-source projects. According to Replit's figures, with GPT-6 Astra serving as the main agent that directs the task (which Replit calls the core loop), Replit Agent scored 72% at a cost of $2.11 per task. Astra running alone at low effort in mini-swe-agent, the basic setup used on the public leaderboard, scored 67% at a cost of $1.60. These are test results published by Replit itself. Replit also said its design lets the model decide how hard to think, when to delegate work and to whom.
According to TechCrunch, Replit sells a cloud software development platform where users can build and deploy apps directly in the browser; under its Anthropic partnership, Replit ran the platform on Claude models. Masad said enterprise made up about 2% of revenue last year, and enterprise contracts make up about 15% of revenue this year. He also said more than 50% of revenue comes from B2B, meaning sales to business customers. The company has not disclosed its ARR, or annual recurring revenue.
What it means for ordinary readers
The reports do not say how this news specifically affects Replit users. If the platform changes, rely on Replit's official announcements. This article is not product-purchase advice.
Frequently asked questions
Has Replit really ended its partnership with Anthropic?
According to TechCrunch, Masad said so on stage, but the same report cites an anonymous source who says the partnership is still active, and TechCrunch has asked Replit for clarification. It is not independently confirmed.
What reasons did Masad give?
TechCrunch quotes him as saying the decision was based on cost, partnership and performance, and that Sol and Astra do better at building vibe-coded apps from scratch.
Why do outsiders have doubts?
When Replit announced its OpenAI partnership in August, there was no sign it had stopped working with Anthropic. And a September blog post still listed Claude Fable 5 and Fable 5.1 as models running in Agent production.
Do I need to change how I use Replit?
The reports do not say how users are specifically affected. Watch for official announcements from Replit.
Browse the latest news in this topic
Lifestyle
Anthropic Pledges 5% of Its Shares to a Public Benefit Fund: What We Know, the Limits, and Open Questions
Anthropic says that, with shareholder approval, it will commit 5% of its shares to a new Public Benefit Fund. This piece covers the company's statements, the fund's structure, how it differs from the existing $200 million research fund, and what it means for readers.
Lifestyle
Anthropic Discloses Internal Shifts: Fewer Junior Technical Roles, More Time in Meetings
AI company Anthropic has published research on employment at its own firm. It says the share of junior staff and new hires in technical teams has fallen sharply, AI agents now work more hours than employees, and employees spend more time in meetings. The report shows job seekers and workers how a heavy AI user is changing hiring and work, but it reflects only that one company.
Lifestyle
Meta Removes Disputed Commercial Terms Language Read as Allowing Scraping of News Sites for AI Training
The Verge reports that Meta, after criticism, revised its commercial terms to remove wording critics read as letting it scrape content from news and other sites to train AI, even overriding site blocks. Meta says the terms apply only to its ads and business products, and site owners can still state preferences via robots.txt.
Lifestyle
Websites Are Blocking AI Agents: Why New Booking and Shopping Features Often Fail
The Verge and TechCrunch report that Amazon, Yelp, Delta Air Lines and other sites are restricting AI agents that act on users' behalf. Some blocks are deliberate; others are anti-bot checks catching agents by mistake. Here is what each side says and what it means for users.