Anthropic Blocks Claude Subscriptions for Third-Party AI Tools

Anthropic has introduced a new policy blocking Claude Pro and Max subscribers from using their subscriptions with external AI platforms like OpenClaw and similar agentic tools. The change took effect on April 4, 2026.

According to Anthropic, the reason is that its subscription models were not built for the usage patterns of third-party tools, and such usage was placing an "outsized strain" on the company's compute and engineering resources.

Users who wish to continue using Claude models through external platforms must now switch to per-token "extra usage" billing or direct API access. For heavy users, this could mean a cost increase of 10 to 50 times their previous monthly expenses.

To ease the transition, Anthropic is offering a one-time credit equal to the monthly plan price, redeemable until April 17, 2026. Discounts of up to 30 percent are available on pre-purchased usage bundles, and full refunds are an option.

OpenClaw was specifically named as an affected tool, but Anthropic has indicated the restriction will extend to all third-party harnesses in the coming weeks. Anthropic's own products — Claude.ai and Claude Code — remain accessible under existing subscription plans.

The decision has sparked debate about the boundaries between subscriptions and API access, and who ultimately controls access to frontier models.