Claude Code is temporarily doubling its usage limits

Anthropic doubles Claude Code usage limits during off-peak hours. Discover how to take advantage, exact schedules, and eligible plans.
•
Updated on August 14, 2026
•
12 min read
Believemy logo

Since March 13, 2026, Anthropic has been doubling Claude's usage limits (including Claude Code) during off-peak hours, through March 27, 2026. This automatic promotion applies to Free, Pro, Max, and Team plans, allowing you to send twice as many messages without any action on your part. Here's everything you need to know to make the most of it.

Good to know

To learn even more, check out our complete Claude Code course.

 

Why Is Anthropic Doubling Claude Code's Limits?

To understand this promotion, you need to put it in context. In early March 2026, Claude experienced an unprecedented wave of popularity.

Following Anthropic's refusal to remove certain ethical safeguards from its AI models for the U.S. Department of Defense, the company was listed as a "supply chain risk" and lost its federal contract.

In response, many users boycotted ChatGPT — whose maker OpenAI had signed a deal with the Pentagon — and migrated to Claude.

The result was spectacular: Claude became the #1 free app on the App Store, dethroning ChatGPT for the first time. Even singer Katy Perry publicly mentioned using Claude. Faced with this massive influx of users, Anthropic had to tackle two simultaneous challenges:

  • Managing server load: peak usage during U.S. business hours put the platform under pressure
  • Retaining newcomers: many new users quickly hit usage limits, which are notoriously strict on Claude

The off-peak limit doubling promotion elegantly addresses both problems: it rewards users while encouraging them to shift their usage to low-activity periods, thereby smoothing out the load on infrastructure.

 

Promotion Details: Dates, Schedules, and Eligible Plans

Promotion Period

The promotion runs from March 13, 2026 to March 27, 2026 at 11:59 PM PT (Pacific Time). That's exactly two weeks to take advantage of the doubled limits.

 

Off-Peak Hours

The doubled limits apply outside of the following peak window, on weekdays:

Time ZonePeak Hours (normal limits)Off-Peak Hours (doubled limits)
Pacific Time (PT)5:00 AM – 11:00 AM11:00 AM – 5:00 AM (next day)
Eastern Time (ET)8:00 AM – 2:00 PM2:00 PM – 8:00 AM (next day)
GMT / London12:00 PM – 6:00 PM6:00 PM – 12:00 PM (next day)
CET / Paris-Berlin1:00 PM – 7:00 PM7:00 PM – 1:00 PM (next day)
AEST / Sydney11:00 PM – 5:00 AM5:00 AM – 11:00 PM
Good to know

On weekends, doubled limits apply all day, 24/7. It's the ideal time to launch intensive Claude Code sessions!

 

Eligible Plans

The promotion covers the following subscriptions:

  • Free — the free plan also benefits from doubling
  • Pro ($20/month)
  • Max 5x ($100/month)
  • Max 20x ($200/month)
  • Team

Enterprise plans are excluded from this promotion. However, Enterprise users generally have much higher token budgets and are not subject to the same constraints.

 

Tools Covered

The doubled limits apply across the entire Claude ecosystem:

  • Claude (web, desktop app, mobile)
  • Claude Code (the command-line interface for coding)
  • Claude Cowork
  • Claude for Excel
  • Claude for PowerPoint

 

How the Doubling Works in Practice for Claude Code

If you use Claude Code daily, here's what this promotion concretely changes for you.

The Five-Hour Sliding Window

Claude Code operates on a five-hour sliding window system. This means your usage limits progressively reset over a five-hour period. During the promotion, the number of messages and the amount of tokens you can consume within this window are doubled.

To illustrate: if your plan normally allows 50 messages per five-hour window, you get 100 during off-peak hours. If you have a token budget of X tokens per window, it goes up to 2X.

 

Checking Your Usage in Real Time

To track your token consumption with Claude Code, you can use the /stats command directly in your session:

BASH
/stats session

This command displays your remaining quota and your window reset time. It's the most reliable way to know whether you're currently in the doubled-limit period or not.

 

Strategies to Maximize the Doubled Limits

Getting double the tokens is good. Knowing how to use them optimally is better. Here are concrete strategies tailored to developers who use Claude Code.

1. Schedule Your Intensive Coding Sessions on Weekends

On weekends, doubled limits apply 24/7. It's the ideal time to:

  • Launch agentic loops (with ralph loops) on major refactors (we cover this in our Claude Code course)
  • Run complete code reviews across your entire codebase
  • Experiment with subsystems or new architectures

 

2. Shift Your Agentic Workflows to Off-Peak Hours

If you're in France or Europe, off-peak hours start at 7:00 PM (CET) on weekdays: perfect for launching long tasks in the evening. You can configure Claude Code to run automated tasks with the following command:

HTML
claude-code review --scope=full --output=report.md

The idea is to prepare your prompts and instructions during the day, then launch execution in the evening when limits are doubled.

 

3. Leverage Multiple Tabs

With double the available tokens, an effective strategy is to open multiple Claude Code sessions in parallel. Instead of waiting for a response from one session, you can work on another aspect of your project in a second terminal:

HTML
# Terminal 1: refactoring the authentication module
claude "Refactor the auth module to add Facebook auth"

# Terminal 2: meanwhile, security review
claude "Run a security audit on the API module"

 

4. Use Claude Code Skills to Automate Your Reviews

If you've configured skills (custom capabilities) in Claude Code, the promotion is the ideal time to chain them together. You can create an orchestrator skill that successively invokes multiple review skills:

YAML
# Example orchestrator skill
name: full-audit
description: Runs all reviews in sequence
steps:
  - invoke: architecture-review
  - invoke: security-review  
  - invoke: performance-review
  - invoke: code-quality-review
output: Unified report with results from each review

This type of workflow consumes a lot of tokens under normal circumstances. With doubled limits, it becomes perfectly viable.

 

What This Promotion Reveals About the Future of AI Pricing

Beyond the promotion itself, this off-peak limit doubling foreshadows major developments in AI service pricing.

Toward Dynamic Time-Based Pricing

Several analysts and developers in the community have noted the similarity to electricity pricing: peak / off-peak rates. This is no coincidence. AI inference is directly tied to GPU energy consumption, and energy costs vary considerably depending on the time of day.

It's very likely that in the future we'll see:

  • Off-peak plans at reduced prices (e.g., $5–10/month for off-peak-only access)
  • More expensive tokens during peak hours and cheaper ones at night
  • Queue systems for non-urgent workloads, similar to the batch APIs already offered by OpenAI

 

Demand Shaping as a Long-Term Strategy

This promotion is also a real-world test of « demand shaping »: the ability to influence user behavior to smooth out load. If Anthropic finds that users actually shift a significant portion of their usage to off-peak hours, this could become a permanent feature of Claude plans.

Good to know

As one Hacker News commenter noted: « It's faster to change user behavior than to buy and install new hardware. » Anthropic is clearly playing both sides: investing in infrastructure AND optimizing the distribution of existing load.

 

The Psychological Aspect: The Abundance Effect

There's also an important psychological dimension. Several users reported that when they have tokens in abundance, their behavior changes dramatically:

  • They open multiple sessions in parallel
  • They explore "side quests" without hesitation
  • They use Claude for complete reviews of their codebase
  • They stop counting their tokens and focus on their project

A Max 20x plan user described the transition as « going from a capped data plan to an unlimited one ». The return to normal limits is then perceived as painful, which can encourage upgrading to a higher plan. It's a subtle but effective marketing strategy.

 

Comparison with Competing Offerings

This promotion comes in a context of fierce competition in the AI assistant market. Here's how the main competitors stack up:

ServiceBase LimitsRecent PromotionsNotes
Claude Code (Anthropic)Varies by plan2x during off-peak (March 2026)Opus 4.6 recognized as the best coding model
Codex (OpenAI)Frequent limit resetsRecently relaxed limits2x more permissive according to some users
Gemini CLI (Google)Seemed nearly unlimitedQuotas recently reduced with Gemini 3.1 ProStrict weekly limits appeared
Copilot (Microsoft)Generous, multi-model—Access to Claude, OpenAI, and MS models
Good to know

An interesting point: while Claude is boosting its off-peak limits, Google seems to be going in the opposite direction, recently imposing stricter weekly quotas on Gemini 3.1 Pro. Several users reported going from a feeling of nearly nonexistent limits to exhausting their quota in just two hours of use.

The most popular strategy among experienced developers remains maintaining subscriptions with 2–3 different providers (typically Claude + a competitor at $20/month each) to diversify their token sources and never be blocked.

 

Limitations and Points to Watch

The Promotion Has an End Date

The doubled limits end on March 27, 2026 at 11:59 PM PT. After that date, all limits return to their standard levels. No changes are made to your plan or billing.

 

Peak Hours Remain Unchanged

During peak hours, your limits remain exactly the same as before the promotion. Don't plan your big agentic coding sessions during this window.

 

Impact on Response Quality

Several users noted a degradation in Opus response speed during peak hours, with speeds sometimes dropping to 2–3 tokens per second on the web. One developer even nicknamed this phenomenon the « Americans waking up syndrome » (observed around 3–4 PM CEST). If response quality is crucial for your workflow, off-peak hours are doubly advantageous: more tokens AND faster responses.

 

The Daylight Saving Time (DST) Trap

The promotion period (March 13–27) partially overlaps with daylight saving time changes in several regions. The United States already changed clocks in early March, but Europe doesn't change until late March. Make sure to verify the correspondence for your time zone by consulting the off-peak hours table earlier in this article.

 

Precedent: The 2025 Holiday Season Promotion

This isn't the first time Anthropic has offered this type of promotion. From December 25 to 31, 2025, a similar doubling was offered, but only to Pro, Max 5x, and Max 20x subscribers. The March 2026 promotion is therefore more generous since it also includes Free and Team plans, and lasts two weeks instead of one.

Good to know

The December promotion had a notable impact on conversions: several users reported discovering the full extent of Claude's capabilities by using more tokens, which encouraged them to upgrade to higher plans. A Reddit user admitted: « The December promotion made me switch to the $100 plan, and I'm not the only one. »

Anthropic has clearly learned from this experience and broadened the scope of the March promotion to reach an even wider audience, particularly free users discovering Claude following the Pentagon controversy.

 

Practical Guide: How to Get the Most Out of It

Here's an actionable summary to make the most of this promotion:

Step 1: Check your current plan. Make sure you're on a Free, Pro, Max, or Team plan. Enterprise plans are not eligible.

Step 2: Identify your local off-peak hours using the time zone table above.

Step 3: Schedule your most token-hungry tasks (code reviews, refactors, architectural explorations) during off-peak hours or on weekends.

Step 4: Monitor your consumption with the /stats session command in Claude Code to verify that the doubled limits are active.

Step 5: Prepare your prompts and instructions in advance during peak hours, then launch execution during off-peak hours.

HTML
# Check your current quota and reset time
/stats session

# Launch an intensive task during off-peak hours
claude "Analyze the complete project architecture and generate a detailed report with improvement recommendations"
Good to know

Reminder: you don't need to do anything to activate the promotion. The doubling is automatically applied as soon as you use Claude outside of peak hours.

 

Frequently Asked Questions

Question

Do I need to do anything to activate the doubled Claude Code limits?

No, absolutely nothing. The promotion applies automatically to all eligible users (Free, Pro, Max, and Team plans). As soon as you use Claude Code outside of peak hours (8 AM – 2 PM ET on weekdays), your limits are automatically doubled. No promo code, no settings to change in your account.

 

Question

Does the bonus usage count toward my weekly limit?

No. The additional usage obtained during off-peak hours is not charged against your weekly limits. Concretely, your "normal" usage (the first half of the five-hour window) is deducted normally from the weekly quota, but the bonus portion granted by the promotion is entirely exempt. It's a genuine gift in terms of total capacity.

 

Question

What happens after March 27, 2026?

Usage limits simply return to their standard levels, at all hours of the day. There are no changes to your plan, subscription, or billing. You won't be charged more and you won't lose any features — you'll simply go back to the usual limits of your plan.

 

Question

Does the doubling work on weekends?

Yes, and it's the best time to take advantage of it. On weekends, doubled limits apply 24 hours a day, with no time restrictions. It's therefore the optimal time to launch intensive coding sessions with Claude Code, run complete code reviews, or experiment with new features.

 

Question

Why are Enterprise plans excluded from the promotion?

Anthropic hasn't given an official reason, but Enterprise plans operate on a different pricing and provisioning model, with generally much higher limits and specific service level agreements (SLAs). These accounts have dedicated token budgets that are already sized for intensive use. Additionally, the U.S. government agencies that terminated their contracts with Anthropic were on Enterprise plans — it's possible the exclusion is also related to this context.

Share this article

Want to help us? Share this article on your networks or even better: on your site, in an article or in your newsletter.

Believemy logo
Comments (0)

You need to be logged in to comment on this article: log in or sign up.

Read next

Claude is increasing its quotas by 25%, and yet you are going to lose some.

Anthropic raises Claude weekly limits by 25% on September 14. Why your real quota can still drop despite the increase, and how to adapt your workflow.

August 30, 2026

Claude Code and Elon Musk: The SpaceX Deal Changing Everything

Anthropic signs with Elon Musk's SpaceX for 220,000 Nvidia GPUs. Discover the impact on Claude Code, doubled limits, and what it means for developers.

May 9, 2026

Claude Pro, Max, Team: what you actually get over the free plan

Free, Pro, Max 5x, Max 20x, Team: real limits per 5-hour window, prices, access to Fable 5.1 and Claude Code, and the point where paying more pays off.

September 4, 2026

Claude Opus 4.8: Features, Benchmarks & Complete Guide

Anthropic's Claude Opus 4.8 focuses on honesty, parallel agents, and more reliable code. Discover everything that changes and how to make the most of it.

May 29, 2026

Claude Opus 5: Anthropic's New Flagship AI Model

Discover Claude Opus 5: performance, pricing, benchmarks and use cases. Anthropic's most capable AI model for coding and analysis.

July 24, 2026

Claude Sonnet 5: Anthropic's Most Agentic AI Model

Discover Claude Sonnet 5, Anthropic's most agentic AI model: performance, pricing, use cases and API access. Everything you need to get started.

June 30, 2026

Claude Max API credits: how to claim your $200 every month

Your Claude Max plan hides $100 to $200 of API credits a month, but you have to claim them. We show you where to click, and the trap to avoid when you confirm.

October 8, 2026

7 Claude Code mods to create when you work alone (no coding)

Tempted by mods but not sure what to ask for? Here are 7 ideas that help when you work alone, each with the sentence to paste into Claude Code.

October 13, 2026

Become the solopreneur you dream of being.

First project, career change, business launch: access our courses, practice on real-world cases, and move forward with clear guidance.