How does Claude Code wait for a usage limit to reset and continue the task on its own?

Claude Code is Anthropic's coding tool. You type a request in a terminal, the text window where you type commands, and an AI model reads files, runs commands and edits code for you. On a Pro, Max, Team or Enterprise subscription, that work counts against a usage limit: an allowance of work the plan includes. The work is done in tokens, the pieces of text the model reads and writes, and Anthropic states the allowance as a multiple of another plan's allowance, with no token count attached. The allowance has a session window that resets every five hours and a weekly window, both explained on the page about the five-hour window and the weekly window. When a window runs out in the middle of a task, Claude Code stops. This page is about what you can do at that moment, and in particular about one feature: Claude Code can keep the session open, wait, and continue the task on its own when the limit resets. It is part of the guide to Claude Code usage limits.

Every fact on this page comes from Anthropic's own documentation and support articles, read on 4 October 2026. Anthropic's figures are its own statements about its own product.

The message you see when a limit stops you

Anthropic's error reference lists four messages for a subscription plan's allowance running out:

You've hit your session limit · resets 3:45pm
You've hit your weekly limit · resets Mon 12:00am
You've hit your Opus limit · resets 3:45pm
You've hit your Sonnet limit · resets 3:45pm

Each message shows the reset time, and Anthropic states that "Claude Code blocks further requests until the reset time shown in the message" [3]. The first two limits cover every model, so switching models does not restore access. The Opus and Sonnet limits each apply to one model family, and switching to a model outside that family with /model keeps you working [3]. The error reference adds that each model has its own prompt cache, the stored copy of a request that Anthropic's service reuses to avoid processing the same text twice, so the first request after a switch re-reads the whole conversation with no cache hits [3]. The page on switching model or effort mid-task covers that cost.

What the wait feature does

Anthropic's interactive mode documentation, read on 4 October 2026, describes the feature in one sentence: "When a claude.ai usage limit stops Claude mid-task, Claude Code waits in the open session and continues the task on its own after the limit resets." Automatic continue is on by default in interactive sessions signed in with a claude.ai subscription, and the feature requires Claude Code v2.1.234 or later [1]. The error reference states the same version: before v2.1.234, Claude Code did not offer the wait [3].

While Claude Code waits, two lines at the bottom of the session show when the limit resets and when Claude will continue [1]:

Usage limit reached · limit resets 3:45pm
Continuing automatically at 3:45pm · esc to cancel

Either line can have more text after those words, such as a help link on the first line or /usage-credits to continue now on the second. When the wait starts on its own, the conversation also records a line reading Usage limit reached · continuing automatically at 3:45pm · esc to cancel [1].

Anthropic's instruction is to keep the session open. What happens next depends on how the wait ends [1]:

• At the reset. The second line changes to Continuing shortly · esc to cancel. Then Usage limit reset · continuing automatically appears in the conversation, and Claude Code prompts Claude to pick the task up where it stopped. It does not resend your last message.
• After your computer slept. If the computer slept for longer than a period Anthropic puts at 30 minutes, and the limit reset while it slept, the first line reads Your usage limit has reset and the second reads Press enter to continue. You press Enter. After a shorter sleep, or one that ended before the reset, Claude Code continues on its own.
• Early. When you finish adding usage credits with /usage-credits, sign back in after /upgrade, or switch models with /model during the wait, Claude Code checks whether usage is available again and continues right away if it is. It does not check after an upgrade or purchase you make in a browser on your own. Under opusplan, a model setting that runs plan mode (the mode in which Claude proposes an approach before editing anything) on a different model from the one used for the work, and under other model settings that run plan mode on a different model, Claude Code waits for the reset instead.

The continued task runs like any other turn. Claude Code still asks for permissions, the approvals it requests before an action such as running a command, so the task can stop on a prompt while you are away. If the continued task reaches the limit again, Claude Code restarts the wait on its own at most twice in a row, then stops and shows Automatic continue stopped after repeated usage-limit hits · /rate-limit-options to try again [1].

When Claude Code starts the wait on its own, and when you start it

The documentation separates three situations: when Claude Code starts the wait on its own, when you start it yourself, and when it is never offered.

Claude Code does not start the wait on its own in these cases [1]:

• Remote Control and agent team teammate sessions. Remote Control is Anthropic's feature for continuing a session on your computer from a phone, tablet or browser. An agent team is a group of Claude Code sessions working together, and a teammate is one of them. In both cases, a person at that terminal can still start a wait.
• A reset more than 24 hours away. Anthropic's example is a weekly limit, which "can reset days out".
• An Opus or Sonnet limit while you run a model outside that family. Your next turn may not reach that limit. opusplan and other model settings that run plan mode on the limited family do not get this exception.

In those cases, and whenever automatic continue is off, Claude Code opens the usage-limit options menu once per reset window when you reach a limit at your own terminal. You pick the row that starts Wait here, then continue automatically to start the wait. In a Remote Control or agent team teammate session, you run /rate-limit-options yourself to open the menu [1].

Claude Code does not offer the wait at all in these cases [1]:

• Background sessions and -p runs. A -p run is a non-interactive run started from a script with the -p flag. The menu row is not available there. The guide to running Claude Code from a script with a spend cap covers how such runs are controlled.
• API keys, cloud providers, and usage-based billing. An API key is the credential for paying Anthropic per token through the Claude Console, Anthropic's developer account site. Anthropic's reason is that usage there "is metered per request, so there is no reset to wait for". Usage credits are the paid use that continues after the included limit; when you finish adding them with /usage-credits during a wait, Claude Code checks whether usage is available and continues right away [1]. The page on usage credits after the limit explains them.
• An LLM gateway without a saved claude.ai login. An LLM gateway is a server placed between Claude Code and the model provider, which forwards requests. Claude Code offers the wait only while a saved claude.ai login is the active credential.

How a wait ends without continuing

Besides a cancel, the documentation lists five ways a wait ends with no continuation [1]:

• You send a prompt. Claude Code sends your prompt instead of waiting. If your prompt reaches the limit too, it stays in the conversation and Claude Code starts the wait again.
• You exit Claude Code. The wait does not restart when you resume the session.
• The conversation changes hands. You switch accounts with /login, clear or rewind the conversation, /resume another session, pull one with /teleport, relaunch with /tui, or hand the session to Claude Desktop, a background session, or the cloud.
• The setting turns off, or the reset moves past 24 hours. This ends only a wait Claude Code started on its own. A wait you picked from /rate-limit-options keeps counting down.
• The continuation is blocked. A hook is a script of yours that Claude Code runs at a set moment; a UserPromptSubmit hook runs when a prompt is submitted. If such a hook blocks the continuation prompt, or a failure stops it before it reaches the model, the wait ends and Claude Code tells you the continuation did not run. You send a prompt to continue.

How to cancel the wait

Press Esc at an empty prompt, or Ctrl+C, while the two lines show, or run /rate-limit-options and pick Don't continue automatically. Claude Code confirms with a line that starts Automatic continue cancelled [1].

After a cancel, nothing continues until you send a prompt or pick Wait here, then continue automatically from /rate-limit-options again. Claude Code does not start a wait on its own again for that reset window; the next reset window starts fresh [1].

The setting autoContinueAtUsageLimit

To turn automatic continue off, open /config and turn off Continue automatically at usage limit, or set autoContinueAtUsageLimit to false in your user settings. /config autoContinueAtUsageLimit=false also works, including with -p. The key=value form cannot turn the setting back on, and Anthropic gives the reason: "the setting grants unattended execution" [1].

Anthropic's settings reference, read on 4 October 2026, gives the key's rules [2]:

• Scope. User or managed. It is read from user settings, --settings, and managed settings only. Managed settings are the settings an administrator applies to every machine in an organisation. When none of those three sets the key, a project or local settings file that sets it turns the feature off instead of being ignored.
• Type and default. Boolean, default true. With true, Claude Code waits in the open session and continues the task after the reset. With false, Claude Code does not start the wait on its own, and you can still start one yourself from the usage-limit options menu.
• In /config. The row Continue automatically at usage limit writes this key to user settings, and Claude Code hides the row while managed settings or the --settings flag set the key.

{
  "autoContinueAtUsageLimit": false
}

Anthropic's cost documentation adds the organisation's side: an administrator who wants to control for a whole fleet whether Claude Code starts the wait on its own sets autoContinueAtUsageLimit in managed settings [4]. The guide for teams covers that in settings an administrator can enforce, and the spend-limit messages an administrator sees are on what each limit message means.

The Desktop app has its own checkbox

In the Code tab of the Claude Desktop app, the session-limit card offers an Auto-continue when limits reset checkbox. The weekly-limit card does not. When it is checked, the Desktop app retries the interrupted turn after the reset and shows the retry time on the card. Anthropic states that the Desktop checkbox and the CLI's Continue automatically at usage limit setting in /config are separate, so you turn each off on its own [3].

The other options when a limit stops you

Anthropic's support article on using Claude Code with a Pro or Max plan, read on 4 October 2026, lists the options for a subscriber who reaches a limit [5]. For a Pro subscriber: consider upgrading to Max 5x if you reach limits consistently, enable usage credits to continue after the included limit, switch to a Claude Console account and buy API usage for periods of intensive coding work, or wait until the limits reset. For a Max 5x subscriber, the upgrade step points to Max 20x, and the other three options are the same [5]. The article also says that to stay inside the plan you decline the API credit option when it is presented and let the usage period reset [5].

The error reference adds /usage to see your plan limits and when they reset, /usage-credits to buy additional usage on Pro and Max or request it from an administrator on Team and Enterprise, and the pricing page for a plan upgrade [3]. Which plan gives how much is on the page about what each plan includes, and the choice between a subscription and an API key is on subscription or API key for one developer.

The options compared

Option	What happens	When it applies
Automatic wait	Claude Code keeps the session open and continues the task on its own after the reset; at most two automatic restarts in a row [1]	Interactive session, claude.ai subscription, v2.1.234 or later, reset within 24 hours, not a Remote Control or teammate session, a limit your current model would reach, and the setting on [1]
Wait you start yourself	Pick Wait here, then continue automatically from /rate-limit-options; the wait keeps counting down even past 24 hours or after the setting turns off [1]	Any interactive subscription session where the wait is offered, including Remote Control and teammate sessions and weekly resets [1]
Wait with no automation	Do nothing until the reset time in the message; Claude Code blocks requests until then [3]	Every plan and sign-in method, including -p runs [1][3]
Switch model	/model to a model outside the limited family keeps you working, at the cost of one uncached request [3]	Opus or Sonnet limit only; the session and weekly limits cover all models [3]
Usage credits	Continue at standard API rates after the included limit; /usage-credits opens the settings [3][5]	Pro and Max directly; Team and Enterprise through an administrator [3]
Upgrade	Pro to Max 5x, or Max 5x to Max 20x [5]	A subscriber who reaches limits consistently [5]
Console account	Buy API usage for periods of intensive coding work [5]	A subscriber who needs more than the plan at that moment; billed per token, no wait offered [1][5]
Desktop app checkbox	The app retries the interrupted turn after the reset [3]	Session-limit card in the Desktop app's Code tab only [3]

Which option to pick

Reveneau is an AI software development consultancy whose code is written by AI, so a usage limit in the middle of a task is a cost of every Reveneau build, and Reveneau is independent of Anthropic. Its recommendation from these facts: leave automatic continue on for a session limit, because the reset is within hours and the session continues without you typing anything. For a weekly limit, which Anthropic says can reset days out, the wait suits it less well: start it yourself only if the task can wait that long, and otherwise decide between usage credits and waiting, using the page on usage credits after the limit. Turn the setting off on a machine where a task continuing without a person present would be a problem, because Anthropic itself describes the setting as granting unattended execution [1]. Every rule above is as Anthropic published it on 4 October 2026, and the page on the dated log of limit changes records changes to the limits themselves.