Claude Code does not have one reset time for every user. For individual Pro and Max subscriptions, Anthropic documents a session-based usage limit that resets every five hours and a separate weekly limit that resets at a fixed day and time assigned to the account. The time shown for your account is the one to trust—not a schedule copied from somebody else's account.
Start with the exact limit message, confirm how Claude Code is signed in, then check Settings → Usage in Claude or run /usage in Claude Code. If /usage says it is showing last-known usage, retry before treating its timer or percentage as current.
The short answer depends on the limit
“Usage reset” can describe several different events. They do not have the same clock or remedy.
| What stopped you | What resets or changes | What to check |
|---|---|---|
| Pro or Max session-based usage | The documented session allowance resets every five hours | The limit message, Settings → Usage, or /usage |
| Pro or Max weekly usage | The weekly allowance resets at the fixed day and time assigned to your account | Settings → Usage for your account's next reset |
| Conversation context or length | Starting, clearing, or compacting a conversation changes the conversation state | The context warning, not the plan-usage timer |
| Enterprise included usage | The applicable pool and rolling-window message depend on the seat and organization plan | Your sign-in path, limit message, and administrator policy |
| API-key usage | Usage is billed separately instead of consuming the same included subscription allowance | API billing, rate limits, spend limits, and provider settings |
This is why “Claude Code resets every five hours” is useful but incomplete. It describes the session cadence Anthropic documents for personal Pro and Max subscriptions; it does not tell you whether a weekly cap, organization policy, context limit, or separately billed API path is blocking your work.
Find your actual reset time in two places
For a personal Pro or Max account, open Claude → Settings → Usage. Anthropic's plan documentation identifies this as the place to see the next reset time for the five-hour and weekly limits. That value is account-specific, so a public article cannot calculate it for you.
Inside Claude Code, run:
text/usage
The usage view can show plan bars and activity. Read its status as well as its percentages: the Claude Code cost documentation says that when the usage request is rate-limited, /usage may show a last-known snapshot from the previous 60 minutes and offer a retry. If the timer or utilization looks wrong, retry and compare it with Settings → Usage before concluding that the reset is late.
Also read the original limit message closely. A message about a weekly limit is not superseded by a newly available five-hour window. A model-specific limit may leave another supported model available, while a context warning points to the current conversation rather than the subscription allowance.
Diagnose the limit before you wait
Use this sequence when Claude Code stops mid-task:
- Identify how Claude Code is authenticated. A Claude Pro or Max sign-in, an organization seat, and an API key do not share one metering contract.
- Name the limit shown in the message. Look for session-based, weekly, model, context, organization, API rate-limit, or billing language.
- Check a current account surface. Use Settings → Usage or
/usage; retry if the CLI labels its data as last known. - Match the remedy to the limit. Wait for the displayed reset, manage the conversation context, ask an organization administrator, or inspect API billing and rate limits as appropriate.
- Treat unexplained behavior as unresolved. If the authenticated surfaces disagree after a refresh, check for a current service incident or contact Anthropic support with the exact message, time zone, plan, and sign-in path.
That final step matters when the limit appears to reset early, late, or more than once. Community reports can reveal an anomaly worth investigating, but an observation from another account does not replace your account's current state or the current plan contract.
/clear does not reset plan usage
The /clear command starts a new conversation and resets local session totals. It can be useful when the current conversation is no longer needed, but Anthropic's documentation does not describe it as resetting included Pro, Max, Team, or Enterprise plan usage.
The same distinction applies to other context-management actions:
- Starting a new conversation gives you a fresh context, not fresh subscription usage.
- Compacting a conversation reduces the context burden; it does not establish that a five-hour or weekly allowance has reset.
- Logging out and back in can change which account or authentication path you are using, but it is not a documented quota-reset control.
If a fresh conversation works when the old one did not, you may have resolved a context or length limit. That is a real fix, but it is different from resetting time-based account usage.
Can you manually reset Claude Code usage?
The checked Anthropic documentation does not identify a supported user control that immediately resets included Claude Code plan usage on demand. That is an evidence-bounded statement, not a claim that an internal, support-led, incident-specific, experimental, or future mechanism can never exist.
Be especially careful with “reset hacks” based on sending a scheduled throwaway prompt. Even if a user observes a different window boundary afterward, the checked first-party documentation does not establish that technique as an on-demand quota reset. It may also consume usage at a time you did not intend.
Likewise, an unexpected early weekly drop in measured utilization does not prove a new universal reset schedule. Plan changes, promotions, incidents, experiments, and display freshness can produce observations that need account- and date-specific verification.
What can you do after reaching the limit?
For Pro and Max users, Anthropic documents several supported paths: wait for the applicable reset, enable usage credits where available, change plans, or choose separately billed API usage. Availability depends on the account and can also vary by region, organization policy, and Claude Code version.
These choices are not interchangeable:
- Waiting preserves the current included plan and resumes when the displayed allowance resets.
- Usage credits can extend work where the account offers them, but they add usage-based charges.
- Changing plans changes the allowance and cost; it is not an instant reset guarantee.
- API usage is a separate pay-as-you-go billing path. It may still have rate limits, spend limits, budgets, or cloud-provider restrictions, and it does not reset your subscription allowance.
Before switching authentication just to continue a task, verify where charges will go and whether your organization permits the new path.
Why the weekly reset may not match someone else's
For individual Pro and Max plans, Anthropic says the weekly limit resets at a fixed day and time assigned to the account. Another user's weekday, hour, or time zone is therefore not a reliable calculator for yours.
Team and Enterprise arrangements require an additional check. Included-usage seats and usage-based organization plans do not share one universal reset contract, and administrator settings can affect what is available. If Claude Code is authenticated through an organization, confirm the seat type and policy instead of importing personal Pro or Max rules.
The same caution applies when a cloud provider or API key is involved. A pay-as-you-go path may not produce the subscription hard stop you expected, but it can stop for its own billing, rate-limit, budget, or provider reasons.
A reliable answer is the one your account can verify
If you use individual Pro or Max, remember the documented baseline: five-hour session-based usage plus a separate account-assigned weekly reset. Then check Settings → Usage or a fresh /usage response for the time that applies to you.
If the message, sign-in path, and usage surface point to something else, do not force that baseline onto the problem. Separate plan usage from conversation context, organization metering, API billing, and temporary anomalies. The useful answer is not a clever universal clock; it is the limit that actually stopped your work and the next step your account can verify.
