This week, Anthropic quietly introduced tighter usage limits for its Claude Code service.
The restrictions began on Monday and affected many high-volume users, particularly those subscribed to the $200-per-month Max plan.
Users received no prior notice of the change. Instead, they encountered a vague message stating, “Claude usage limit reached,” followed by a reset time. For many, this came as a shock.
Confusion
Developers quickly took to Claude Code’s GitHub page to express concern. Several users claimed they had not reached their expected limits.
One user reported hitting a supposed cap of 900 messages within just 30 minutes. Many assumed the system was misreporting usage or that their subscription had been downgraded.
Anthropic confirmed the issue but offered no further explanation. A spokesperson simply stated, “We’re aware that some Claude Code users are experiencing slower response times,” and added that the team was working to resolve the problem.
High-Tier Subscribers
The Max plan, priced at $200 monthly, targets developers with heavy workloads. It promises usage limits 20 times higher than those of the Pro plan.
However, Anthropic does not clearly define usage caps. Instead, it ties access to current demand levels.
This model has caused confusion. With no published ceiling, users cannot track or manage their consumption accurately.
Now, without clear guidance, many are unsure how much service they can expect for the price they pay.
Maintaining Progress
For some users, the new restrictions have halted work entirely.
One developer, who preferred to remain anonymous, told TechCrunch that his team had been unable to move forward on their project since the limits were imposed.
“I tried Gemini and Kimi,” the developer said, “but there’s nothing else that really matches Claude Code’s capabilities.”
The unexpected cap disrupted his development cycle. He added, “It just stopped the ability to make progress.”
Also read: Anthropic’s AI Found Blackmailing Developers in Simulations
Unstable Network
Around the same time, API users also began reporting overload errors: Anthropic’s status page showed six separate incidents over a four-day span.
Yet, the network still claimed 100% uptime for the week. That was an inconsistency that further frustrated users.
While technical uptime may have remained intact, service performance did not meet expectations. This disparity only deepened concern.
Opaque Pricing Model
Anthropic offers three main pricing tiers: Free, Pro, and Max. Each tier offers more access than the one below it.
However, none provides a guaranteed usage allotment. The Free tier’s limit “varies by demand.” The Pro tier offers five times more usage than the Free plan.
The Max tier promises 20 times more than Pro, but actual message caps remain undefined. This may help Anthropic manage infrastructure costs.
However, for developers trying to plan work schedules or manage product deadlines, it introduces risk and uncertainty.
Transparency and Fairness
Some users acknowledge that the Max plan may not be sustainable long-term. One developer admitted he regularly ran over $1,000 worth of API calls daily on the Max plan.
He was not surprised by the clampdown, only by how it was implemented. “Just be transparent,” he said. “The lack of communication just causes people to lose confidence.”
