Claude Code's usage limits go permanent on 14 September — and for heavy users it's a real cut

Anthropic is replacing Claude Code's temporary 50% weekly usage boost with a permanent 25% increase over its pre-May baseline from 14 September 2026 — a change Anthropic itself confirms works out to roughly a 17% reduction for anyone currently running on the boosted allowance.

Anthropic is making Claude Code’s weekly usage limits permanent on 14 September 2026, and the framing depends entirely on which number you compare against. Measured against the pre-May 2026 baseline, the new limit is a 25% increase — Anthropic’s headline. Measured against what heavy users have actually been running on since the temporary 50% boost kicked in, it’s roughly a 17% cut, which Anthropic has confirmed itself. Both figures are accurate; they’re just answering different questions. The temporary boost stays live through 13 September, then the new permanent limits apply across Pro, Max, Team, and seat-based Enterprise plans.

This lands squarely in a pattern search interest has been tracking all year: teams and agencies that built delivery workflows around Claude Code’s usage ceiling — running multiple long agentic sessions per week, per developer — have had to keep re-checking what “unlimited enough” actually means as Anthropic adjusts the dial. It’s the same dynamic we flagged when the temporary boost was extended through 19 August: usage limits on AI coding tools aren’t a fixed cost input the way a seat licence is. They move, and the direction isn’t always up.

Why this matters if you’re commissioning, not just coding

If you’re paying an agency or in-house team that leans on Claude Code for a chunk of delivery, this is worth a direct question before 14 September, not after: does a 17% cut to their current throughput change your timeline or their capacity, and do they have a fallback if it does? A team dependent on a single AI coding vendor’s usage tier has less room to absorb a change like this than one that routes work across tools based on the task.

So what

This is a small mechanical change with a real operational consequence for teams that budget delivery capacity against AI tool usage ceilings. A development partner worth using treats that ceiling as one input among several, not a hard constraint that reshapes what gets built and when. See how we approach AI-assisted development without over-indexing on any single vendor’s terms, or get in touch if a usage change like this is affecting a project you’ve commissioned.

← All research

More research.

GitHub Copilot SDK is becoming the Microsoft Copilot SDK: what builders of AI products should know

On 28 October 2026 GitHub said its Copilot SDK is being renamed Microsoft Copilot SDK, the programmable layer for the shared Microsoft and GitHub agent harness. No immediate action is needed.

Read ›

GitHub Agent Merge: Copilot now watches your pull requests and fixes what blocks the merge

On 28 October 2026 GitHub added Agent Merge to pull requests: a 'Watch and fix' option where Copilot repairs failing Actions workflows and merge conflicts in the cloud, optionally with auto-merge.

Read ›

Claude Code 2.1.314 adds a setting that blocks AI file writes outside the project folder

Claude Code 2.1.314 (30 October 2026) adds permissions.blockWritesOutsideWorkingDirectories, so Claude's file-editing tools refuse any path outside the project in every permission mode.

Read ›

GitHub Enterprise now includes Copilot: seats are bundled, AI usage moves to your budget

From 28 October 2026 account-team customers can get Copilot included with every GitHub Enterprise licence, with AI usage metered against an enterprise budget from the first request. Standalone Copilot is $10 per user per month.

Read ›

GitHub Copilot Fusion goes GA: one pick, many models, and a 30% launch discount

On 28 October 2026 GitHub made Copilot Fusion generally available: instead of choosing a model, developers pick an effort level and Copilot routes each task across models. Paid plans get 30% off model costs for a limited time.

Read ›

GitHub Agentic Workflows is generally available: automations written in plain Markdown

On 28 October 2026 GitHub made Agentic Workflows generally available: AI agents that triage issues, analyse CI failures and update docs, defined in Markdown and run inside GitHub Actions with read-only defaults.

Read ›

Start a project

Building with AI? Let's talk.

Tell us what you are building. We'll give you an honest view of the right approach.

Start a projectSee our work