Anthropic PBC, the developer and provider of the Claude artificial intelligence model family, said it’s offering a special credit pool for users who want to use agentic tools with its large language models. 

The move Wednesday comes as centralized AI vendors come to terms with how they want to handle agentic AI for coding and personal use as adoption rises. Examples include open-source projects such as OpenClaw and Hermes, which were built to provide frameworks for AI agents that can act as 24/7 personal assistants and connect to LLMs such as OpenAI Group PBC’s GPT models or Claude. 

Wednesday, on the ClaudeDev X account, Anthropic said that every paid tier, beginning June 15, will gain a special “programmatic credit pool” that will refill every month and allow users to draw token credits for agentic use. 

Starting at $20 per month for the lowest tier and going up to $200 for Max 20x, this will allow users to connect their third-party agentic tools to Claude and use it as a “brain” until their credits run out. Afterward, they can continue to pay for extra usage to continue what the company calls “interactive” use; users who do not have this enabled will have their usage cut off. 

It also separates the agentic use case from the subscription chat credits, which power the everyday usage, such as talking to the AI models, summarizing documents, using Cowork to sort through emails and other business or personal tasks on the platform. 

Originally, AI agents such as OpenClaw connected through the main subscription and did not need credits. However, last month, Boris Cherny, head of Claude Code at Anthropic, announced that the company was cutting off access to third-party agent frameworks via the subscription and directed users to the Anthropic API, which runs on a per-credit basis. 

This move led to steeper monthly costs for users. At the time, Cherny said that the rise of OpenClaw usage had put strain on Anthropic’s servers because of greatly increased demand and that subscription Claude accounts were not permitted to use AI agents in the first place. 

That change was the first attempt by Anthropic to isolate “interactive” uses from subscription use. 

The subscription model meets agentic compute

Shortly thereafter, the company experimented with removing Claude Code, the company’s agentic coding capability, from a small percentage of new Pro signups. It was a test that the company quickly pulled back from after headlines and consumer backlash. 

Although the credits change opens up OpenClaw users to Claude once again, not all users find Anthropic’s move benevolent. Many have painted the change as a regression, which will make the subscription model worse for existing users. 

“For everyone running real automation, this is a downgrade dressed up as a feature,” wrote arckollect on X. 

In particular, users noted that those using agentic tools would burn quickly through the new cap and that the credits would not roll over month to month when unused. This means that the power users the cap is aimed at could simply hit it within days, while those who use agentic tools dynamically month to month could run out suddenly one month and find credits completely wasted in another. 

Anthropic is not alone in trying to put a meter around agentic work. GitHub recently announced it’s moving Copilot to AI Credits on June 1, and infrastructure providers such as Fireworks AI Inc. and Together Computer Inc. already price serverless inference by token usage rather than by flat subscription. Although Fireworks has been exploring a flat rate subscription called Fire Pass with a monthly token offering within it to scoop up those “prosumer” agentic users. 

The difference is that Anthropic is making this shift inside a consumer-facing subscription product, where power users had already built habits, tools and expectations around Claude as a persistent agentic “brain.” 

The reason behind this is that agentic usage represents great computational draw, multiple rapid tool calls and high reasoning demand. The number of tokens used escalates quickly while an agent “thinks” through a problem, making flat rates less lucrative for providers. 

For enterprises, the shift also signals where agentic AI costs are likely headed. Personal subscriptions may be useful for experimentation, but production agents that run continuously, call tools and reason across long workflows may increasingly need to live inside rationed environments with budget controls, monitoring and predictable limits. Anthropic’s move puts a clearer wall between casual use and automated compute, and that wall is likely to become more common as AI model and inference suppliers figure out how to price for agents that never sleep. 

Images: Anthropic, X