AI coding · fixes · 2026-10-23

Claude weekly limit reached? Why it burns fast + what to do

As of 24 October 2026. Anthropic changes plans, limits, and messages from time to time, so check the linked help pages before you rely on any step.

This guide pairs with Ship-Ready AI Rules, and if you want the next step, read Gemini API key reported as leaked? Rotate, cap, find the leak.

Claude Code "weekly limit reached": why it runs out fast and what to do

Anthropic's help centre says that when you reach a limit without usage credits turned on, you'll see "You've reached your weekly limit. It resets [day] at [time]." (Troubleshoot Claude error messages). Claude Code's own docs word it as "You've hit your weekly limit" (Manage costs effectively). Either way, it's mid-week and you still have work to ship. Here's what to do now, why the week drained so fast, and how to make next week last longer.

Quick fix: what to do right now

1. Check when it resets. The error itself shows the reset day and time. Anthropic suggests you "monitor your remaining allocation using the /status command" (Use Claude Code with your Pro or Max plan), and on a paid plan /usage shows plan usage bars and a breakdown (costs docs).
2. Don't expect a model switch to help. The docs say a weekly limit is "shared across all models, so the developer can't restore access by switching models with /model." Switching models only helps after a model-specific message such as "You've hit your Opus limit" (costs docs).
3. Decide whether to pay to keep going. With usage credits on, your requests "switch to pay-as-you-go pricing on usage credits" and you'll see "Now using credits. Your plan limit resets [day] at [time]" (help centre). We cover the settings in Claude Code monthly spend limits and usage credits.
4. Or finish in another tool. Write a short handoff note (goal, files touched, what's left, how to test) and carry on elsewhere. See the fallback plan below.
5. Or wait. Anthropic lists "Wait until your usage limits reset" as an option for both Pro and Max (help centre).

Why your weekly limit is draining faster than last week

Anthropic says Pro and Max plans "have a five-hour session limit and a weekly limit", that Max plans "also have a separate weekly limit for Fable", and that "these limits are shared across Claude and Claude Code" (help centre). So chat use counts too. Inside Claude Code, these are the usual causes.

Long sessions and big context

The docs explain that "Claude Code sends your full conversation with every request", so "a one-line question in a session that has been open all day still draws usage for the whole conversation". Your first message after a long break can also miss the cache and reprocess your full context (Why usage climbs in a long session). Use /clear between unrelated tasks. Note that /compact "reads the conversation it summarizes, so compacting a large context is itself a large request", while /clear "costs nothing".

Failed command loops and autonomous runs

Users have reported both. One issue describes a session that "burned far more tokens than the task needed" through repeated failed shell commands and retries (#104474, 23 Oct 2026). Another describes an autonomous multi-agent run that "consumed most of my weekly Claude subscription usage" over three days (#104501, 23 Oct 2026). The docs add that subagents, agent teammates, and scheduled tasks each send their own requests (costs docs).

A heavier model on simple tasks

The docs say "Sonnet handles most coding tasks well and costs less than Opus", and suggest model: haiku for simple subagent tasks. Lowering effort with /effort is another lever (costs docs).

Reported metering changes

One Max user reports their weekly meter went "0% to 54% in 22h with same token volume that took 4.4 days to reach 60% the prior week" (#103376, opened 19 Oct 2026, open when we checked). That is a user report, not a confirmed change. If your usage pattern didn't change but the meter did, add your data to the issue and contact Anthropic support.

How to check where the week went

• /usage on a paid plan shows "a breakdown of what counts against your plan limits, with recommendations", including the share used by skills, subagents, plugins, and MCP servers. Press d or w for the last 24 hours or 7 days. The docs note these figures are approximate and "usage from other devices or claude.ai is not included" (costs docs).
• A status line can show usage continuously, and /context shows what's taking up space (costs docs).

Settings designed to make next week last

• Set a smaller default model and a smaller subagent model for routine work.
• Keep CLAUDE.md short. The docs suggest aiming for "under 200 lines" and moving workflow-specific instructions into skills.
• Send noisy searches, test runs, and log reading to a subagent so only a summary returns.

We cover each of these step by step in how to reduce Claude Code token usage, so we won't repeat them here. To estimate API costs if you move to pay-as-you-go, try the AI API cost calculator.

Fallback plan when the limit hits

If you also have Codex, keep a project AGENTS.md with the same rules as your CLAUDE.md, write a handoff note, and continue there until the reset. Codex has its own limits and credits, which we explain in Codex credits, costs, and usage limit resets and ChatGPT Pro and Codex usage limits.

FAQ

When does the Claude Code weekly limit reset?

The limit message shows the day and time it resets, and /status and /usage show your current position. We don't list fixed reset times because they're shown per account.

Is the 5-hour limit different from the weekly limit?

Yes. Anthropic describes them as two separate limits on Pro and Max, with different messages: "5-hour limit reached - resets [time]" versus the weekly message (help centre).

Does Max 20x have a weekly limit?

Anthropic says both Pro and Max plans have a weekly limit, and Max plans also have a separate weekly limit for Fable. It suggests Max 5x users who consistently hit limits consider Max 20x (help centre). We don't quote limit sizes because Anthropic's help pages we checked don't publish them as fixed numbers.

Why did my limit drain faster this week?

Most often: longer sessions, more subagents or automation, failed command loops, or a heavier model. Some users report a possible metering change (#103376). That isn't confirmed.

Can I buy more usage?

Yes, through usage credits on paid plans, or by switching to a Claude Console account with API credits (help centre).

Does /compact save usage?

It can shrink later requests, but the compaction itself "reads the conversation it summarizes", so it's a large request on a large context. If you don't need the history, /clear costs nothing (costs docs).

Sources checked 24 Oct 2026: support.claude.com articles 12466728 and 11145838; code.claude.com/docs/en/costs; github.com/anthropics/claude-code issues #103376, #104474, and #104501.

Claude and Claude Code are trademarks of Anthropic PBC. Codex and ChatGPT are trademarks of OpenAI. GitHub is a trademark of GitHub, Inc. Stack & Ship is not affiliated with or endorsed by them.