Anthropic lifts OpenClaw ban, introduces Agent SDK credits

Anthropic has reversed its ban on OpenClaw, allowing users to once again utilize Claude subscriptions for third-party AI agents through a newly introduced Agent SDK credits tier. This policy change, announced in May 2026, comes after a brief period where OpenClaw users faced restrictions due to concerns over runaway capacity costs associated with third-party agent loops. The new structure provides a separate metered pool of credits for agent usage, ensuring that users can effectively manage their API consumption alongside their conversational quotas.

With the introduction of the Agent SDK credits, OpenClaw users can now access Claude Pro and Max subscriptions while benefiting from a pricing model that aligns closely with their usage needs. The credits are designed to facilitate AI automation and enhance agent workflows, providing developers with the necessary tooling to integrate OpenClaw seamlessly into their applications. This update not only restores access but also offers a more sustainable approach to managing resources, allowing for continued innovation in the OpenClaw ecosystem.