Anthropic has ended the complimentary access to its Claude AI for integrations in external applications. In a post on X, Boris Cherny, who leads Claude Code at Anthropic and is one of its creators, announced that standard Claude subscriptions will no longer include free usage of the AI agent within third-party software, including tools like OpenClaw. Starting at 3 p.m. ET on April 4, individuals accessing Claude via such apps or programs must now purchase an additional usage package or obtain a Claude API key, as stated by Cherny.
While much of Claude's activity involves basic queries from users, a segment of its users interacts with the AI via OpenClaw, an open-source and cost-free AI helper developed by the creator of Moltbook. Distinct from broader AI platforms, OpenClaw focuses on streamlining individual routines, such as managing email backlogs, dispatching messages, or arranging schedules, and it relies on outside large language models like Claude, ChatGPT, and Google Gemini.
Responding to inquiries on X, Cherny attributed the policy shift to technical limitations and efficiency efforts. He noted on the platform that the team has been striving to handle the surge in Claude's popularity, but the existing subscription model was not designed to accommodate the distinct consumption habits of third-party integrations. Cherny emphasized careful resource allocation for capacity and a focus on direct customers of Anthropic's offerings and API.
Those continuing to employ OpenClaw with Anthropic's language model will need to acquire a usage package, which is presently available at a reduced rate, or opt for alternatives such as xAI, Perplexity, or DeepSeek. Anthropic provides its own comparable solution, Claude Cowork, which addresses many of the same automation functions as OpenClaw.