Claude Code usage limit reached: what to do

The direct answer

Below: every message Claude Code shows, a decision tree, a worked example, the errors that look like a limit but aren’t, and what to change so it happens less often. If you only want the time, our Claude reset time tool turns the message into a countdown in your time zone.

The exact messages, and what each one means

Claude Code’s error reference lists four usage-limit messages for people signed in with a Pro or Max plan. The time and day in yours will differ:

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

Two details matter. First, session and weekly limits apply across every model, and Anthropic says the limits are shared across Claude and Claude Code. A long chat on claude.ai uses the same allowance as your terminal. Second, Max plans have a separate weekly limit for Fable. When it runs out, Claude Code can continue on usage credits, but only after you confirm. If the prompt went unanswered, you’ll see something like this and nothing was sent:

Fable limit reached · continuing on Fable 5.1 uses usage credits, and the prompt
to confirm went unanswered — nothing was sent · answer it where this session is
running, or /model to change

Decision tree: what to do right now

Work down the steps. Most people are unblocked by step 3.

Option 1: let Claude Code wait for you

Since v2.1.234, when a claude.ai usage limit stops Claude mid-task, Claude Code waits in the open session and picks the task up on its own after the reset. It’s on by default in interactive sessions signed in with a claude.ai subscription. While it waits, the bottom of the session shows two lines:

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

At the reset you’ll see Usage limit reset · continuing automatically, and Claude Code prompts Claude to continue where it stopped. It doesn’t resend your last message. A few things the docs call out:

• Keep the session open. If you exit, the wait doesn’t restart when you resume.
• Sleep is mostly fine. If your Mac slept more than about 30 minutes past the reset, you’ll see Press enter to continue instead.
• Permissions still apply. The continued task can stop on a permission prompt while you’re away.
• It gives up after repeats. If it hits the limit again, it re-arms at most twice in a row, then shows Automatic continue stopped after repeated usage-limit hits.
• Weekly limits usually need a manual start. Claude Code doesn’t start the wait on its own when the reset is more than 24 hours away. Pick Wait here, then continue automatically from /rate-limit-options if you want it anyway.
• It’s not offered everywhere. Background sessions, -p runs, API keys and cloud providers don’t get the wait, because there’s no plan reset to wait for.

To cancel, press Esc at an empty prompt or pick Don’t continue automatically in /rate-limit-options. To turn the feature off, switch off Continue automatically at usage limit in /config. In the Desktop app’s Code tab, the equivalent is the Auto-continue when limits reset checkbox; the two settings are separate.

Option 2: keep going on usage credits

On Pro, Max 5x and Max 20x you can turn on usage credits to keep working after you reach your included limits. In Claude Code, run /usage-credits; on Pro and Max it opens Settings → Usage on claude.ai, where you enable credits, add funds and optionally turn on auto-reload. Anthropic’s help center says usage credits are billed at standard API rates, separately from your subscription, and apply to Claude Code terminal usage too.

Set a monthly spend limit when you enable them. If you reach it, Claude Code shows You've hit your monthly spend limit · raise it at https://claude.ai/settings/usage… and, on Pro and Max, offers to raise or remove the cap without leaving the CLI. If you finish adding credits while Claude Code is waiting for a reset, it checks again and continues straight away.

Two gotchas. The monthly API credits that come with Max can’t be used in Claude Code or the Claude apps. And once you’re drawing on credits, Claude Code drops the main conversation’s prompt cache from one hour to five minutes, so long pauses get more expensive. /usage shows a usage-credits row with this month’s spend while credits are on.

Option 3: a free limit reset, if you have one

Anthropic occasionally gives eligible plans a one-time limit reset. If you have one, Settings → Usage on the web or in Claude Desktop shows a Reset for free button in the Resets section, and the same button can appear on the limit message in those apps. It isn’t available in Claude Code in a terminal or IDE, so open claude.ai to use it.

Depending on the offer, it restores either your five-hour session limit or your weekly limit. It can’t be undone, it expires at the time listed, and your weekly limit still resets on its usual day and time afterwards.

Option 4: switch tools, or switch models

For an Opus or Sonnet limit, /model is the whole fix. That limit covers only one family, so a model outside it keeps working. Pick the model at a natural break if you can: each model has its own prompt cache, so the first turn after a switch reprocesses the whole conversation.

For a session or weekly limit, a different model won’t help, but a different tool will. If you also have a ChatGPT plan, Codex has its own separate allowance (see Claude Code vs Codex limits). Commit or stash first and write a short handoff note so the other agent doesn’t start cold.

One switch to avoid by accident: if ANTHROPIC_API_KEY is set in your shell, Claude Code uses it instead of your subscription, and you pay API rates. Environment variables take precedence over /login. Run /status and check the API key row if you’re unsure.

Errors that look like a limit but aren’t

These messages come from capacity or credentials, not your plan allowance. Waiting for your plan reset won’t fix them, and they don’t mean you’ve used anything up:

Could it be a bug?

Sometimes the display is the problem, not the allowance. A few things to check before assuming you’re out:

• Update Claude Code. Run claude --version. The docs list several fixed usage-display issues, such as an MCP server’s share of usage being overstated before v2.1.222, and /usage showing no bars on a rate-limited request before v2.1.208.
• Look for “Showing last-known usage”. If the usage request itself is rate limited, /usage shows bars it loaded in the past 60 minutes. Press r to retry.
• Remember the breakdown is local. The /usage breakdown is estimated from session history on this machine, so usage on claude.ai or another computer won’t appear in it, but it still counts.
• Check status.claude.com for incidents.
• Billing questions go to support: on claude.ai, click your initials, then Get help.

Make it happen less often

Claude Code’s cost docs explain why a session drains faster than it feels like it should. Each request resends the whole conversation, so a one-line question in a session that’s been open all day still draws usage for all of it. The fixes are mostly habits:

1. /clear between unrelated tasks. Clearing costs nothing; /compact is itself a large request.
2. Mind the one-hour cache. On a subscription the main conversation’s cache lasts an hour. Your first message after a longer break reprocesses everything.
3. Match the model to the job. The docs suggest Sonnet for most coding and Opus for complex reasoning. Subagents can run on Haiku.
4. Watch subagents, teams and loops. Each subagent sends its own requests against the same limits. Scheduled tasks fire while you’re idle.
5. Check the breakdown. On paid plans, /usage flags behaviors that account for 10% or more of recent usage, such as long context or cache misses.

And watch the number before the wall. You’ll get warning messages as you approach a limit, the status line can show your five-hour and weekly percentages (here’s how), and Quotchi keeps both in the Mac menu bar with alerts at the levels you choose.

Common mistakes and myths

“Switching to Sonnet gets me past the session limit.”

It doesn’t. Session and weekly limits apply across all models. Switching only helps with an Opus-specific limit.

“The session resets five hours after I started.”

Maybe, maybe not. Anthropic doesn’t document what starts the clock. Use the time in the message or /usage.

“Restarting Claude Code clears the limit.”

No. The limit is on your account, shared with claude.ai and Desktop. Restarting also cancels a pending auto-continue.

“Max’s monthly API credits will cover me.”

Anthropic says they can’t be used in Claude Code. Usage credits are a separate thing.

“Support can reset it.”

The help center describes occasional one-time resets that you use yourself from Settings → Usage, not resets on request.

“A 529 means I’m out.”

No. It’s capacity, and the docs say it doesn’t count against your quota.

What changed recently

Sources

Official pages we read on October 9, 2026. If anything here disagrees with them, they win.

• Claude Code docs: Errors (usage limit messages)
• Claude Code docs: Interactive mode (wait for a usage limit to reset)
• Claude Code docs: Manage costs effectively
• Claude Code docs: How Claude Code uses prompt caching
• Claude Help Center: Use Claude Code with your Pro or Max plan
• Claude Help Center: How do usage and length limits work?
• Claude Help Center: Manage usage credits for paid Claude plans
• Claude Help Center: What is a limit reset?
• Claude Help Center: What is the Max plan?

FAQ