Anthropic is cutting Claude Code's current weekly limits by 17%

What happened

Anthropic announced this week that it is permanently raising the standard weekly usage limits for Claude Code by 25 percent across its Pro, Max, Team, and seat-based Enterprise plans. On the surface, that sounds like a win for developers who rely on the AI coding assistant for heavy workloads. In practice, the new permanent baseline sits below the temporary elevated limits users have been operating under, meaning the practical experience for many will feel like a cut rather than a raise.

This piece breaks down exactly what changed in Claude Code's limit structure, how the math leads to a net 17 percent reduction for users currently at the top of the temporary ceiling, and what developers and engineering teams on each affected plan should think about before their next billing cycle. If you run Claude Code in a professional or enterprise context, this change is relevant to your workflow today.

Anthropic announced a permanent 25 percent increase to Claude Code's standard weekly usage limits for subscribers on the Pro, Max, Team, and seat-based Enterprise plans. The company framed this as an expansion of access to the AI coding tool. However, because Anthropic had previously been running those limits at a higher temporary level, the permanent number lands roughly 17 percent below where the current ceiling actually sits. Users who built workflows around the elevated temporary limits will find themselves with less headroom than they have now, even though the official framing emphasizes an increase over the original baseline.

How it works

The announcement did not include a specific date for when the rollback takes effect beyond the implication that it is imminent and permanent. The 25 percent figure applies to the standard limit tier, not to any add-on or overage capacity that individual plans may support separately.

Claude Code tracks usage against a weekly rolling limit rather than a monthly cap. When Anthropic introduced elevated temporary limits, it effectively let that ceiling float above the published standard. The permanent 25 percent increase brings the official standard limit up, but because the temporary ceiling had been set even higher to accommodate demand or testing, the resulting permanent number is lower than what users currently experience. The arithmetic is straightforward: if the original standard was 100 units, the new permanent standard is 125 units, but the temporary ceiling was sitting at approximately 150 units, making the effective change a reduction of roughly 17 percent from where developers are operating today.

Why it matters

Builders verifying their own exposure should check which plan tier they are on — Pro, Max, Team, or seat-based Enterprise — since each has its own absolute limit values even if the percentage adjustments move in parallel. The mechanism for tracking and enforcing these limits remains the same weekly-reset model already in place.

Claude Code has become a primary interface for developers using Anthropic's models inside editors and automated pipelines. Weekly limits are not an abstract policy detail — they directly constrain how much autonomous coding, review, and generation a team can run before hitting a wall and either waiting for the reset or paying for additional capacity. A 17 percent reduction from the current working ceiling is meaningful for any team that regularly approaches its limit, because those teams by definition have calibrated their workflows around the higher number.

The framing of this change as an increase is technically accurate relative to the original published standard, but it obscures the real operational impact for users who have been living inside the elevated temporary window. Transparency about the delta from the current limit, not just from the historical baseline, would give engineers a clearer picture of what to plan for.

Who is affected

Every Claude Code subscriber on a Pro, Max, Team, or seat-based Enterprise plan is subject to the new permanent limit. Individual developers on Pro plans who use Claude Code for personal projects may notice the ceiling only if their usage routinely runs high. Max plan subscribers, who pay more for expanded access, are likely the cohort most sensitive to a 17 percent reduction because they opted into a higher tier specifically to avoid hitting limits. Team and seat-based Enterprise customers are affected at an organizational level, where multiple engineers drawing from a shared or per-seat limit can collectively feel the constraint faster than any single user would.

Volume-API Enterprise plans, if structured differently from seat-based Enterprise, may not be covered by the same adjustment — developers on those arrangements should verify their specific contract terms rather than assume the same percentage changes apply.

What to watch next

The most immediate thing builders should do is audit their current weekly Claude Code usage against the new permanent limit to identify whether they will hit the ceiling sooner than before. Teams that consistently approach the current temporary limit have roughly a 17 percent buffer reduction to account for starting at the next reset cycle. Anthropic has not indicated whether it will restore or further raise limits in the near term, so planning around the new permanent number is the prudent baseline.

Longer term, the pattern of setting a high temporary limit and then locking in a lower permanent one is worth tracking as a signal of how Anthropic manages capacity and demand for Claude Code specifically. If limit adjustments follow product updates or model releases, watching Anthropic's changelog and plan documentation will give early indication of the next change before it lands in a billing cycle.