Claude Code Desktop Adds Auto-Continue After Usage Limits

Claude Code waits in the open session and continues your task automatically when a claude.ai usage limit resets, and you do not have to switch it on. Anthropic's Interactive mode reference: "Automatic continue is on by default in interactive sessions signed in with a claude.ai subscription. Requires Claude Code v2.1.234 or later." The opt-out is in /config under "Continue automatically at usage limit", or the autoContinueAtUsageLimit settings key. (Interactive mode, Settings reference)

The feature shipped in the changelog for version 2.1.234 as a single line ("Claude Code now continues your session automatically when a claude.ai usage limit resets; turn it off in /config"). Anthropic's docs now cover the mechanics that line left out: what the wait looks like, what ends it, when Claude Code will not start one on its own, and how an admin turns it off for a fleet. (Claude Code changelog)

Key Takeaways

• On by default in interactive sessions signed in with a claude.ai subscription, from Claude Code 2.1.234.
• Turn it off in /config ("Continue automatically at usage limit") or set autoContinueAtUsageLimit to false. Managed settings can set it for a whole organization.
• At the reset, Claude Code sends a fixed prompt to pick the task up. It does not resend your last message.
• It resumes; it does not add allowance. The limit and the reset schedule are unchanged.
• It is not offered for API keys, cloud providers, usage-based billing, background sessions, or -p runs.
• It does not start on its own in Remote Control or agent-team teammate sessions, when the reset is more than 24 hours away, or for an Opus or Sonnet limit while you run a model outside that family. /rate-limit-options starts the wait by hand.
• The Desktop app has a separate checkbox, "Auto-continue when limits reset", on the session-limit card. Turning off one does not turn off the other.

How Auto-Continue Works

When a claude.ai usage limit stops Claude mid-task, a line at the bottom of the session shows when it will continue: Usage limit reached · continuing automatically at 3:45pm · esc to cancel. Keep the session open. Anthropic documents three ways the wait ends with the task continuing: (Interactive mode)

• At the reset. The line reads continuing shortly, then Usage limit reset · continuing automatically, and "Claude Code sends Claude a fixed prompt to pick the task up where it stopped. It doesn't resend your last message."
• After your computer slept. If it slept for more than about 30 minutes and the limit reset while it slept, the line reads Your usage limit has reset · press enter to continue, and Claude Code waits for you. After a shorter sleep it continues on its own.
• Early. When you add 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.

The continued task runs like any other turn. Claude Code still asks for permissions as usual, so the task can stop on a prompt while you are away. If it hits the limit again, Claude Code re-arms the wait at most twice in a row, then stops and shows Automatic continue stopped after repeated usage-limit hits · /rate-limit-options to try again.

Who Gets It

The default applies to interactive sessions signed in with a claude.ai subscription. Anthropic's Interactive mode reference sorts everything else into three groups. (Interactive mode)

In the cases where the wait does not start on its own, and whenever automatic continue is off, Claude Code opens the usage-limit options menu once per reset window when you hit a limit at your own terminal. Pick the row that starts "Wait here, then continue automatically". In a Remote Control or agent-team teammate session, run /rate-limit-options yourself; the command does not appear in the command menu, so type it in full. (Commands reference)

How to Turn It Off

Three levels, from one wait to a whole organization:

1. Cancel one wait. Press Esc at an empty prompt, or Ctrl+C, while the line shows, or run /rate-limit-options and pick "Don't continue automatically". Claude Code confirms with a line that starts Automatic continue cancelled, and does not start a wait on its own again for that reset window.
2. Turn it off for yourself. In /config, turn off "Continue automatically at usage limit", or add this to your user settings:

{
  "autoContinueAtUsageLimit": false
}

/config autoContinueAtUsageLimit=false also works, including with -p, "but the key=value form can't turn it back on, because the setting grants unattended execution." Turning it back on takes the /config panel. (Interactive mode) 3. Turn it off for a fleet. The key's scope is "User or managed": Claude Code reads it from user settings, --settings, and managed settings only, and hides the /config row while managed settings or --settings set it. A project or local settings file cannot turn the feature on, but if none of the higher scopes sets the key, a project file that sets it turns the feature off. Anthropic's Errors page points admins at managed settings for exactly this. (Settings reference, Errors)

What Ends the Wait Without Continuing

Anthropic lists these: you send a prompt (Claude Code runs it instead of waiting); you exit Claude Code (the wait does not restart when you resume the session); the conversation changes hands (/login to another account, clear or rewind, /resume another session, /teleport, /tui, or handing 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; one you picked from the menu keeps counting down); or a UserPromptSubmit hook blocks the continuation prompt. (Interactive mode)

If you automate around this, the Notification hook has three matchers for it: quota_auto_resume_fired (the task continued), quota_auto_resume_stale (the reset happened during a long sleep and Claude Code is waiting for Enter), and quota_auto_resume_disabled (the wait ended without continuing). The last one does not fire when you press Esc or Ctrl+C or pick "Don't continue automatically". (Hooks reference)

The Desktop Checkbox

Claude Code Desktop has its own control. Anthropic's Errors page: "In the Code tab of the Desktop app, the session-limit card offers an Auto-continue when limits reset checkbox. The weekly-limit card doesn't. When it's checked, the Desktop app retries the interrupted turn after the reset and shows the retry time on the card. The Desktop checkbox and the CLI's Continue automatically at usage limit setting in /config are separate, so turn each off on its own." Anthropic does not state whether the Desktop checkbox is checked by default. (Errors)

What This Does Not Change

Auto-continue is a scheduling convenience, not a new allowance. Anthropic's Help Center says usage limits vary by plan and by factors such as conversation length, features used, and model, and that usage across claude.ai, Claude Code, and Claude Desktop counts toward the same limit. After a limit, the documented options are still to wait for the reset, upgrade your plan, or purchase usage credits; auto-continue automates the first. (Claude Help Center)

Two things to check before leaving a session at a limit overnight. A weekly limit that resets more than 24 hours out will not be waited for automatically; start the wait from /rate-limit-options if you want it. And the continued turn still stops on permission prompts, so a task that needs approvals will wait for you either way.

Corrections and Updates

September 9, 2026: this page said Anthropic documented only the changelog line and did not state plan scope, resume timing, or multi-session behavior. Anthropic's Claude Code docs now cover all of it (the Interactive mode, Settings, Errors, Commands, and Hooks references), and the page is rewritten against them. The August 23 correction below also said the original page's "desktop checkbox" was wrong; Anthropic's Errors page documents an "Auto-continue when limits reset" checkbox in the Desktop app's session-limit card, separate from the CLI setting, so that half of the correction is withdrawn. The CLI default-on finding stands.

August 23, 2026: this page previously described auto-continue as an opt-in desktop checkbox, and both halves of that were wrong.

Sources

• Anthropic, Claude Code docs, "Interactive mode", sections "Wait for a usage limit to reset", "Cancel the wait", "Start a wait yourself", and "Turn automatic continue off": https://code.claude.com/docs/en/interactive-mode#wait-for-a-usage-limit-to-reset (checked September 9, 2026; source for the default, the eligibility conditions, the wait line, the fixed prompt, the sleep and early-continue behavior, the re-arm limit, and every way the wait ends)
• Anthropic, Claude Code docs, "Settings reference", entry autoContinueAtUsageLimit: https://code.claude.com/docs/en/settings-reference#autocontinueatusagelimit (checked September 9, 2026; source for the key, its scope, its default, and the project-file behavior)
• Anthropic, Claude Code docs, "Errors", section "You've hit your session limit": https://code.claude.com/docs/en/errors#youve-hit-your-session-limit (checked September 9, 2026; source for the Desktop checkbox and the managed-settings pointer)
• Anthropic, Claude Code docs, "Commands", rows /rate-limit-options and /config: https://code.claude.com/docs/en/commands (checked September 9, 2026)
• Anthropic, Claude Code docs, "Hooks", Notification matchers quota_auto_resume_fired, quota_auto_resume_stale, quota_auto_resume_disabled: https://code.claude.com/docs/en/hooks#notification (checked September 9, 2026)
• Anthropic, Claude Code changelog, entry for version 2.1.234: https://github.com/anthropics/claude-code/blob/main/CHANGELOG.md (checked September 9, 2026; the release record; the current release is 2.1.266)
• Anthropic Help Center, "How do usage and length limits work?": https://support.claude.com/en/articles/11647753-how-do-usage-and-length-limits-work (checked September 9, 2026; last updated by Anthropic July 13, 2026; source for general usage-limit behavior and the wait, upgrade, or credits options)
• ClaudeDevs on X, August 13, 2026: https://x.com/ClaudeDevs/status/2088014831605702937 (cited as the first public flag of the feature, not relied on; x.com answers automated requests with HTTP 200 and a JavaScript shell containing none of the post text, re-checked September 9, 2026, and cannot be re-verified)