When does a Claude Code usage limit reset? The five-hour window and the weekly window

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 is measured 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. Anthropic's documentation, read on 4 October 2026, describes two windows for that allowance. One resets every five hours. The other resets once a week. This page explains both, where each is documented, what the message says when one runs out, and how to see where you stand. It is the first page of the guide to Claude Code usage limits.

The session window resets every five hours

Anthropic's support article on the Pro plan, read on 4 October 2026, says: "Your session-based usage limit will reset every five hours." The Max plan article uses the same sentence [6][7]. Anthropic's pricing page says every plan has usage limits "that reset on a rolling five-hour session window" [8]. The support article on usage credits, which is what Anthropic calls paid use after the included limit, repeats the figure twice: "Your session limits reset every five hours as usual", and "Your plan's included usage limit will reset every five hours once you reach it" [2].

For Team and Enterprise plans, Anthropic's Claude Code documentation says each member's usage "draws from a per-seat allowance that resets on a rolling five-hour window and a weekly window" [3]. A seat is one member's place on the plan.

The pricing page says what the window holds in relative terms only. Pro "gives you at least 5x more usage per 5-hour session than Free", and Max "gives you 5x or 20x more usage per 5-hour session than Pro" [8]. Across this guide's sources, Anthropic states the allowance only as these multiples, and says the amount of work it allows "depends on the length and complexity of your conversations, the model you choose, and the features you use, so there's no fixed message count" [8]. The page on which plan includes what sets out the multiples by plan.

The weekly window resets at a fixed time assigned to your account

The weekly window is documented in the Pro and Max plan articles, read on 4 October 2026. Both say the plan has "a weekly usage limit that applies across all models", and both describe the reset the same way: "Weekly limits reset at a fixed time each week that is assigned to your account. Your reset day and time stay the same regardless of when you start using Claude or when your subscription begins, and you receive your full weekly allowance each cycle. You can see your next reset time in Settings > Usage" [6][7]. Settings > Usage is a page in your Claude account on claude.ai.

The pricing page states the relationship between the two windows in one sentence: every plan has the five-hour window, "and paid plans add weekly limits" as well [8]. So a Pro or Max subscriber has both windows at once, and Anthropic's documentation for Team and Enterprise seats names both as well [3].

What the word session means here

The word session has two uses in Anthropic's documentation, and the limit messages use the first one.

In the limit messages, session means the five-hour window. The error reference, read on 4 October 2026, shows the message as You've hit your session limit · resets 3:45pm [4]. The Pro and Max articles call the same thing the "session-based usage limit" that "will reset every five hours" [6][7], and the pricing page calls it the "five-hour session window" [8].

In the Claude Code terminal, a session is one conversation. Anthropic's cost documentation says the token totals in the /usage screen "reset when /clear starts a new session" [3]. /clear is the command that empties the conversation and starts a new one. The same screen shows both meanings side by side. Its Session block, Anthropic says, "shows API token usage and is intended for API users", while subscribers "see plan usage bars, activity stats, and a usage breakdown on the same screen" [3]. The plan usage bars are the ones that show the five-hour and weekly windows.

The practical point: the five-hour window belongs to your account, so it keeps running when you start a new conversation with /clear. Every conversation you run, in the terminal, in an IDE extension, or in the Claude app, counts against the same windows, which the page on what counts against a limit covers in full.

The warning before the limit

Anthropic's support article on using Claude Code with a Pro or Max plan says: "To help you monitor your usage, you will see warning messages about remaining capacity" [1]. The error reference gives the shape of one such warning: You've used 85% of your session limit · resets 3:45pm [4]. The usage credits article says you will "see a clear notification when approaching and reaching your included usage limits" [2].

To watch the allowance all the time instead of waiting for a warning, the error reference names two places. In the terminal, you can add the rate_limits fields to a custom status line, which is the line of text Claude Code can show under the conversation. In the Claude Desktop app, you can click the usage ring next to the model picker [4].

The message when a window runs out

The error reference, read on 4 October 2026, lists four messages:

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

Anthropic states that "Claude Code blocks further requests until the reset time shown in the message" [4]. The time in the message is the fact you need. For the session limit, the example shows a clock time on the same day. For the weekly limit, the example shows a day and a time.

The first two messages cover every model. Anthropic writes: "The session and weekly limits are shared across all models, so switching models doesn't restore access" [4]. The last two apply to one model family each, and switching to a model outside that family keeps you working; the page on what counts against a limit covers those two messages.

Since Claude Code v2.1.234, an interactive session signed in with a claude.ai subscription can wait in the open session and continue the interrupted task after the reset [4]. The page on waiting for a reset covers that feature, including the Desktop app's Auto-continue when limits reset checkbox, which Anthropic says is offered on the session-limit card only [4].

Usage counts against both windows at the same time

Anthropic states this directly: "Usage counts against the session and weekly allowances at the same time. A single burst of heavy activity, such as a large workflow fanout, can exhaust the weekly allowance before the session window resets" [4]. A workflow fanout is a run in which Claude Code starts many agents at once, each with its own conversation; the guide to Claude Code agent costs covers what that costs.

This sentence changes how to read the two bars. A session window gives capacity back within hours. The weekly window gives it back on one fixed day each week [6][7]. Reveneau recommends planning around the weekly bar in /usage for that reason: an empty session bar costs you part of a day, and an empty weekly bar can cost you the rest of the week.

The two windows compared

	Session window	Weekly window
Length	Resets every five hours [2][6][7]	Resets once a week [6][7]
When it resets	Anthropic calls it rolling [3][8]; the message shows a clock time [4]	At a fixed day and time assigned to your account, unchanged by when you start using Claude or when the subscription begins [6][7]
Which plans	Every plan, including Free [8]	Paid plans [8]; named for Pro, Max, Team and Enterprise [3][6][7]
Models covered	All models [4]	All models [4][6][7]
Message at the limit	You've hit your session limit · resets 3:45pm [4]	You've hit your weekly limit · resets Mon 12:00am [4]
Where to see the reset time	The message [4], /usage [4], Settings > Usage on claude.ai [6][7]	The same three places
Does usage count against it	Yes, at the same time as the other window [4]	Yes, at the same time as the other window [4]
Automatic wait in the terminal	Offered from v2.1.234 [4]	See the page on waiting for a reset

The /usage plan bars

Run /usage inside Claude Code to see the plan bars and when each window resets [4]. Anthropic's cost documentation, read on 4 October 2026, says subscribers see "plan usage bars, activity stats, and a usage breakdown" there [3]. The breakdown under the bars shows what recent usage went to, with a d or w key to switch between the last 24 hours and the last 7 days; Anthropic says those breakdown figures are computed from local session history on this machine, so usage from other devices or claude.ai is left out of the breakdown [3]. The page on how to read /usage and /context covers every row of that screen.

When the request for your plan limits fails, most often because the usage endpoint is rate limited, /usage shows the last bars it loaded on this machine within the past 60 minutes, with a note reading Showing last-known usage and the age of that data. Press r to retry. Without a snapshot from the past 60 minutes, it reports that the usage endpoint is rate limited and offers the same retry. Before v2.1.208, a rate-limited request in a session that had not loaded usage yet showed the error with no bars [3].

The Pro or Max support article also says to "monitor your remaining allocation using the /status command" [1], and the Pro and Max plan articles point to Settings > Usage on claude.ai for the next reset time [6][7].

One hour of prompt cache while you are inside the plan

The prompt cache is a store, kept by Anthropic's service, of request text it has already processed, so that the next request can reuse it at a lower cost. Anthropic's documentation says the cache lifetime "is an hour on a subscription and drops to five minutes once you're drawing on usage credits" [3], and its prompt caching page gives the same table: one hour for the main conversation on a Claude subscription within plan usage, five minutes on usage credits, an API key or a cloud provider [5]. The page on the cache lifetime of five minutes or one hour explains what that changes.

Other limits Anthropic reserves

The Pro and Max articles each add one sentence beyond the two windows: "to manage capacity and ensure fair access to all users, we may limit your usage in other ways, such as weekly and monthly caps or model and feature usage, at our discretion" [6][7]. Both articles also mention a limit reset, which they describe as something you can use "to reset your five-hour session limit or weekly usage limits back to full", and point to a separate article for details [6][7]. That article is outside this guide's sources, so this page says no more about it.

What to do when a window runs out

The Pro or Max article lists four options, read on 4 October 2026: upgrade to a higher plan if you reach limits consistently, turn on usage credits to continue after the included limit, switch to a Claude Console account and buy API usage for intensive work, or wait until the limits reset [1]. The next pages take each in turn: which plan includes what, usage credits after the limit, a subscription or an API key for one developer, and waiting for a reset.

Reveneau is an AI software development consultancy, and all of its code is written by AI, so usage limits are a running constraint on every Reveneau build. Reveneau is independent of Anthropic. Every figure on this page is Anthropic's own statement about its own product, read on 4 October 2026, and limit rules change: the page on the dated log of limit changes records what has changed and when.