Anthropic calls it a 25% increase — Claude Code users are calling it a 17% cut
From 14 September 2026, Anthropic is permanently raising Claude Code's standard weekly usage limits by 25% above the original baseline for Pro, Max, Team and seat-based Enterprise plans — but because that date also ends the temporary 50% summer boost, most active users will see their real weekly capacity fall by around 17%, and searches comparing "Claude Code limits" before and after the change have spiked as teams try to work out what it means for their budget.
3 September 2026
Anthropic has announced that from 14 September 2026, Claude Code’s standard weekly usage limits rise permanently by 25% over the plan’s original baseline, applying across Pro, Max, Team and seat-based Enterprise tiers. Read as a headline, that’s good news. Read against what’s actually happening on the same date, it isn’t: the 25% is calculated from the original limit, not from the temporary 50% capacity boost Anthropic has had running through the summer. That boost ends on 13 September — the day before the “increase” lands. Net effect for anyone who has been running on the summer allowance: roughly 17% less weekly capacity than they have right now.
This is the kind of change that reads one way in a press release and another way in a usage dashboard, and it’s exactly the sort of detail that gets missed when a team picks an AI coding tool based on the marketing rather than the metering. It matters because Claude Code, Cursor and GitHub Copilot are now genuinely close competitors on capability, and the tiebreaker for a lot of teams is increasingly usage economics — how much sustained agentic coding work a plan actually buys you before you’re paying overage or downgrading model quality mid-sprint.
Why this matters beyond Anthropic’s own users
Every major AI coding vendor is now managing the same tension: heavy agentic usage is expensive to serve, promotional boosts are cheap to hand out and hard to sustain, and the eventual correction always lands as a limit change that customers discover after the fact rather than one they scoped in advance. If you’re commissioning software from a team that leans on AI-assisted development, it’s worth asking how they handle a vendor tightening limits mid-project — whether their workflow is portable across tools, or whether a single provider’s usage caps can quietly become your delivery risk.
So what
A capable delivery partner treats AI coding tools as leverage, not as a dependency that can move the goalposts on a live project. If you want a build that isn’t hostage to one vendor’s next pricing change, see how we approach AI-assisted development, or get in touch to talk through a project.