Every developer knows the specific frustration of hitting a rate limit in the middle of a complex refactor. You are in a flow state, the AI agent is finally grasping the architecture of your legacy codebase, and suddenly the terminal returns a message stating you have exhausted your quota. This friction point often forces a choice between pausing work for hours or upgrading to a more expensive tier mid-sprint. Anthropic is attempting to alleviate this bottleneck for its power users with a targeted capacity expansion.

The Mechanics of the Capacity Boost

Anthropic has launched a limited-time promotion that increases the weekly usage limits for Claude Code by 50 percent. This adjustment is specifically designed to allow developers to generate and modify larger volumes of code without the immediate pressure of hitting a ceiling. The promotion is active from May 13, 2026, through August 31, 2026. During this window, eligible users will find their resource allocation expanded by half over their standard weekly allotment.

The rollout is targeted at specific subscription tiers. Users on the Pro, Max, and Team plans are fully eligible. Additionally, Enterprise plan users who are on legacy seat-based billing will receive the boost. For these accounts, the increase is applied automatically by the system, requiring no manual activation, admin intervention, or application process.

However, not all users are included in this expansion. Those on the Free plan and Enterprise users utilizing consumption-based billing are excluded from the promotion. Because the eligibility depends on the specific billing structure of the seat, users in corporate environments may need to verify their seat type with their organization's administrator to determine if the boost is active.

Distinguishing Agentic Limits from Chat Limits

To understand the impact of this change, one must distinguish between the general Claude ecosystem and the specific Claude Code toolset. This 50 percent increase applies exclusively to Claude Code, the terminal-based AI coding agent. This includes every interface through which Claude Code is invoked, whether it is the primary CLI, IDE extensions, or the desktop and web interfaces specifically powering the Claude Code agent.

Crucially, this does not extend to the general Claude experience. The usage limits for the standard Claude web interface, the mobile app, and Claude Cowork remain unchanged. This distinction highlights Anthropic's strategy to treat the coding agent as a high-intensity tool with different consumption patterns than a general-purpose chatbot. While a chat might involve sporadic queries, a coding agent often performs recursive loops of reading files, writing tests, and executing commands, which consumes tokens at a significantly higher velocity.

Another critical nuance lies in the type of limit being adjusted. The promotion targets the weekly usage limit only. The short-term, 5-hour burst limits remain exactly as they were. This means that while developers have a larger total pool of resources to draw from over seven days, they are still subject to the same hourly throttling to prevent system abuse and ensure stability. The tension here is between total capacity and instantaneous throughput; Anthropic is increasing the former while keeping the latter capped.

Verification of these limits is handled directly within the developer's environment. There is no need to navigate to a web-based billing dashboard to see the change. By executing a simple command in the terminal, developers can see their real-time status.

/usage

Running this command provides an immediate readout of the updated weekly limit and the current amount of resources consumed. This allows developers to audit their usage and plan high-load tasks, such as major library migrations or comprehensive documentation overhauls, around their remaining quota.

Once the promotion concludes on August 31, 2026, all weekly limits will automatically revert to their standard levels. This transition will not trigger any changes in billing, nor will it force users into different plan tiers. The system will simply slide back to the baseline quota without additional charges or manual configuration.

This temporary expansion suggests a period of aggressive user acquisition and testing for the Claude Code agent, giving developers a taste of higher-ceiling productivity before the standard limits return.