Anthropic Ends Flat-Rate Access to Claude for Third-Party AI Tools, and a Developer Dispute Has Followed
Quick Reads
- Affected subscribers are being offered a one-time credit equal to one month’s plan cost, redeemable until 17 April, and discounts of up to 30 per cent on pre-purchased extra usage bundles.
- Anthropic has blocked Claude Pro and Max subscribers from using their monthly plans to run the AI through third-party tools, effective 4 April 2026.
- The change starts with OpenClaw, an open-source AI agent framework with over 135,000 active users, but Anthropic says it will extend to all outside tools shortly.
- Users who want to keep connecting Claude to third-party tools must now pay separately through a new pay-as-you-go billing system on top of their existing subscription.
- Anthropic says the decision is about computing sustainability, but OpenClaw’s creator, who recently joined OpenAI, says the timing tells a different story.
What Chnaged
Claude Pro costs $20 per month. Claude Max runs between $100 and $200 per month. Both plans have until now allowed subscribers to connect Claude to third-party agentic tools software that can perform tasks on a user’s behalf at no additional cost beyond their subscription. That is now over.
Starting 4 April at 12 pm Pacific Time, subscribers can no longer use their Claude subscription limits for third-party tools, including OpenClaw. Users can still connect those tools to their Claude account, but they will require extra usage and a pay-as-you-go option billed separately from the subscription. The policy applies to all third-party harnesses and will be rolled out to more shortly.
To understand why this matters, it helps to know what OpenClaw is. OpenClaw is an open-source AI agent framework, essentially a custom layer built on top of Claude, that lets users create autonomous AI agents capable of browsing the web, reading emails, sending messages on platforms like WhatsApp and Telegram, and completing multi-step tasks without being prompted every step of the way. It was originally released in November 2025 by Austrian developer Peter Steinberger, who built it as a side project to explore what would happen if a large language model were given persistent memory, tool access, and the ability to communicate through everyday messaging apps. The answer was that an enormous number of people wanted exactly that. More than 135,000 OpenClaw instances were estimated to be running at the time of the announcement, and industry analysts noted a price gap of more than five times between what heavy agentic users paid under flat subscriptions and what equivalent usage would cost at standard API rates. Anthropic’s subscription business was, in effect, quietly cross-subsidising a class of usage it had not priced for.
What Anthropic Says
Boris Cherny, head of Claude Code at Anthropic, confirmed the policy on X. “We’ve been working hard to meet the increase in demand for Claude, and our subscriptions weren’t built for the usage patterns of these third-party tools. Capacity is a resource we manage thoughtfully, and we are prioritising our customers using our products and API,” he wrote.
The technical explanation goes deeper than simple cost management. Anthropic’s own tools, particularly Claude Code, its AI coding assistant, and Claude Cowork, its business automation product are engineered to maximise what the company calls “prompt cache hit rates,” a method of reusing previously processed text to dramatically reduce the raw computing power each session requires. Third-party tools like OpenClaw largely bypass this efficiency layer, meaning a single heavy OpenClaw session consumes far more infrastructure than an equivalent session inside Anthropic’s official products.
One industry analyst observed that a single OpenClaw agent running for a full day could burn the equivalent of $1,000 to $5,000 in API costs, while the user was paying a flat monthly subscription fee. “Anthropic was eating that difference on every user who routed through a third-party harness,” the analyst wrote.
Cherny acknowledged the disruption directly. “I know it sucks,” he wrote in response to an affected developer. “Fundamentally, engineering is about tradeoffs, and one of the things we do to serve a lot of customers is optimise the way subscriptions work to serve as many people as possible with the best mode.” He added that Anthropic is offering full refunds to subscribers who decide the new terms no longer work for them.
The Dispute Behind The Decision
The policy has been received with more than frustration it has landed in the middle of a competitive flashpoint. In February 2026, Steinberger announced he was leaving OpenClaw to join OpenAI. Sam Altman publicly stated that Steinberger would drive the next generation of personal agents at the company, and that OpenClaw would be moved to an open-source foundation with OpenAI’s continued support.
Anthropic’s enforcement came weeks after that move. Steinberger and OpenClaw board member Dave Morin said they attempted to negotiate a softer outcome directly with Anthropic, but were only able to delay enforcement by a single week. “Funny how timings match up, first they copy some popular features into their closed harness, then they lock out open source,” Steinberger wrote publicly following the announcement.
Cherny rejected that framing. He said the decision was driven by infrastructure constraints and noted that he personally submitted code improvements to OpenClaw’s own open-source project to improve its cache efficiency, work he did before and separate from the policy change.
Y Combinator president Garry Tan also weighed in: “Anthropic shutting down OpenClaw may turn out to be a strategic blunder, or strategic genius. The OpenClaw community will be the determiner of whether it is A or B.”
