Claude Code usage limits: the weekly and 5-hour caps by plan
Claude Code usage runs on two rolling windows: a 5-hour session cap and a 7-day weekly limit. Both are set by your plan and shared with the Claude app on Pro and Max - Pro carries the smallest weekly allowance, Max 5x multiplies it roughly five times, Max 20x roughly twenty. The 5-hour window starts with your first message of a session and expires five hours later; the weekly window is rolling, so it never resets at a fixed time - usage ages out over seven days. Anthropic publishes caps as weight-based ranges, not fixed prompt counts, and retunes them often; this page tables the current caps per plan and window, the reset mechanics, and every dated change as it lands.
Claude Code on a paid Claude plan runs against two shared caps: a session limit that resets every five hours, and a weekly limit that resets at one fixed time each week that is assigned to your account 1. One tier now skips the short clock: on the Ultra plan, the October 30, 2026 rename of Max 20x, verified accounts have no five-hour limit at all and live on the weekly cap alone
2. The caps belong to your subscription, not to the terminal: usage in Claude Code, its IDE extensions and the Claude apps all drains one shared pool
3. Max and Ultra plans carry one more window on top of the two: a separate weekly limit for Fable models
3
4. Anthropic publishes no prompt counts; usage is weighted by message length, attachments, model choice and effort level, so every "X prompts per week" figure on a third-party page is an estimate, not a spec
5
1. The ladder in numbers since the October 30 rename: Pro at $20 per month, Max at $100 per month, Ultra at $200 per month, with Max buying 5 times the Pro plan's usage on both clocks and Ultra buying 10 times the weekly allowance with no five-hour clock for verified accounts
1
6
2.
The two clocks: one resets in five hours, one on your account's weekly day
The session cap is the short clock: Anthropic's Pro and Max plan pages both state that the session-based usage limit will reset every five hours 1
6. Ultra is the exception: verified Ultra accounts have no five-hour limit and work inside the weekly cap alone, an account still awaiting verification keeps a temporary five-hour limit until Anthropic's automatic check, which weighs account and payment history, clears it, and the tell sits in Settings, Usage, where the absence of a Current session bar means you are verified
2. The weekly cap applies across all models and resets on a fixed day and time assigned to your account; it stays the same regardless of when you start using Claude or when your subscription begins, you receive the full allowance each cycle, and your next reset is visible in Settings, Usage
6.
Two mechanics matter more than most explainers admit. First, every request counts against the session and weekly allowances at the same time, so a single heavy burst, such as a large workflow fanout, can exhaust the weekly cap while the five-hour window still has room in it 7. Second, alongside the two shared caps sit per-model-family limits for Opus and Sonnet: hitting one blocks only that family, so switching families with /model keeps you working, while switching models never restores the shared session or weekly caps
7. Opus also draws meaningfully more of your quota per turn than Sonnet
8. Fable sits outside both arrangements: on Max and Ultra plans the three current Fable models, Fable 5.2, Fable 5.1 and Fable 5, are included in the subscription but capped at 50% of the weekly allowance, which they use up faster than other models, while on Pro they sit outside the plan allowance entirely and bill pay-as-you-go usage credits from the first request; Fable 5.2, launched November 9, 2026, needs Claude Code version 2.1.321 or later
9
10.
What each tier costs per unit of allowance
The plan pages print prices and multiples but never put the two together. Since the October 30 rebrand the multiple that matters is the weekly one, because the plan pages now print it: the Max page promises five times the Pro plan's usage every five hours and every week, and the Ultra page ten times every week 6
2. Divide the monthly price by the weekly multiple: Pro is 20 ÷ 1 = 20, Max is 100 ÷ 5 = 20, and Ultra is 200 ÷ 10 = 20. Every tier on the ladder costs the same $20 per Pro-equivalent unit of weekly allowance, so an upgrade buys volume, never a discount. Two honesty notes. The retired Max 20x pitch was a per-session multiple, 20 times Pro every five hours, while Ultra's 10x is a weekly figure, and the denominators do not compare, so read the rename as a change of printed window rather than a halving anyone can prove
6
2. And what the extra $100 over Max buys on Ultra is a different clock, not a cheaper unit: verified accounts lose the five-hour wall entirely
2.
The ladder:
- Pro, $20 per month: the 1x baseline for both the five-hour and weekly allowances, and Claude Code plus the Claude apps in one subscription
1
3.
- Max, $100 per month (formerly Max 5x): 5 times the Pro plan's usage every five hours and every week, plus priority access to new models
6.
- Ultra, $200 per month (formerly Max 20x): 10 times the Pro plan's weekly usage, no five-hour session limit once your account is verified, and monthly credits for the Claude API
2.
The weekly cap is not a rolling window
If your mental model is usage aging out over rolling seven days, the vendor's own pages say otherwise: weekly limits reset at a fixed time each week assigned to your account, and the full allowance returns at each reset 1. Claude Code's error reference shows the shape in the wild: the weekly-limit message carries a fixed reset stamp, for example "You've hit your weekly limit · resets Mon 12:00am"
7. The genuinely rolling arrangement exists one tier over: Enterprise seats draw from an organization pool that resets on a rolling window
8. Team and Enterprise metering is a different system, and Anthropic's Team and Enterprise guides own that question
8.
Every dated limit change in 2026
Anthropic retunes these caps often enough that a static explainer is expired on arrival. The dated record so far:
- May 6, 2026: Anthropic doubled Claude Code's five-hour rate limits for Pro, Max, Team and seat-based Enterprise plans, and removed the peak-hour reduction that cut Pro and Max allowances at busy times. The same announcement covered a deal to use all compute capacity at SpaceX's Colossus 1 data center, over 220,000 GPUs and more than 300 megawatts
11.
- May 13 to September 13, 2026: weekly limits ran 50% higher as an automatic promotion, in Claude Code only, across the CLI, IDE extensions, desktop and web. The five-hour caps were not affected, and the Claude apps' limits were unchanged
12.
- July 19, 2026: the promotion that let paid plans spend up to 50% of their weekly limit on Claude Fable 5 at no extra cost ended at 11:59 PM PT. Fable 5 remained included on Max plans, still capped at half the weekly allowance, while on Pro plans and standard Team seats it moved to pay-as-you-go usage credits; Fable 5.1 was never part of the promotion
9.
- September 14, 2026: weekly limits settled at 25% above their pre-promotion level, permanently, for Pro, Max, Team and seat-based Enterprise plans
12
13.
- The arithmetic behind the headlines: 1.5 × 0.83 ≈ 1.25. A 17% cut from the promotional level lands at 25% above the pre-May baseline, which is why one event was covered as both an increase and a cut; Anthropic's own clarification called it "a 17% reduction in weekly limits on Claude Code" compared with the promotional level
13.
- October 30, 2026: the ladder was renamed, not repriced. Max 20x became the Ultra plan at the same $200 per month, existing subscribers moved over automatically at no price change, Ultra prints a 10x weekly multiple and drops the five-hour session limit for verified accounts, and the surviving Max plan, formerly Max 5x, now prints its multiple for both clocks, a first for the weekly dimension
2
6.
- November 9, 2026: Fable 5.2 launched alongside Mythos 5.2, the same underlying model made available through verified research programs, and joined the Fable window: all three Fable models are capped at 50% of the weekly allowance on Max and Ultra, Fable 5.2 needs Claude Code 2.1.321 or later, and a limited-time launch offer adds extra Fable usage on eligible plans
10
9.
- November 10, 2026: Anthropic's ClaudeDevs account on X said the company had fixed "a few small issues that used more limits than intended" and gave every Pro, Max, Ultra and Team subscriber one free full reset across all models and limits, usable until December 10, 2026 at 2:00 PM PT
14. The reset is redeemed through the Reset for free button under Settings, Usage rather than inside the Claude Code terminal, and the restored usage applies across the shared account, Claude Code included
15.
As of this mid-November refresh, the September 14 settle remains the last move in the underlying cap quantities: October 30 reshuffled which tier carries them and November 10 refilled them once for everyone, but the allowances per plan have not moved since; the only other limits change still in flight is the wrap-up allowance documented below, rolling out gradually 12
3.
Where to watch for the next one: these cap changes landed on Anthropic's news page, help-center articles and Anthropic's X account 12
13, not in the Claude Code changelog, which tracks tool behavior release by release, most recently version 2.1.325 on November 13, 2026
16.
What hitting a cap looks like, and what is not a cap
Four messages name the four caps, each with its own reset stamp 7:
- "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"
Warnings arrive early, with a message such as "You've used 85% of your session limit · resets 3:45pm". One behavior now softens the five-hour wall itself: if the session limit lands while Claude Code is partway through a response, it may keep working briefly to finish the step in progress instead of stopping mid-step, showing "Usage limit reached · wrapping up" as it does. The wrap-up allowance is small and capped, its usage counts against the weekly limit, and it covers only a response already under way, never a new message sent after the limit. Pro plans get it at most once per weekly period, while Max plans can get it each time a five-hour session limit lands, within the weekly cap; verified Ultra accounts never see it, because they have no five-hour limit to wrap up against, though an Ultra account still carrying its temporary limit is treated like Max until verification clears. It is rolling out gradually, so it may not be on your account yet 3. In an interactive session signed in with a subscription, the CLI can park itself at a cap and pick the work back up shortly after the reset, via the Continue automatically at usage limit setting in /config, and the Desktop app's Code tab offers its own Auto-continue checkbox on the session-limit card
7. The Max and Ultra Fable window has no "You've hit" message of its own: crossing it surfaces a consent prompt before further Fable requests bill usage credits
7. Two errors look like caps but are not: "Server is temporarily limiting requests" is a short-lived throttle unrelated to your quota, and "Request rejected (429)" points at an API-key rate tier, not your subscription
7. "Usage credits required for 1M context" is an entitlement check on the million-token window, not an exhausted quota
7. Run /usage to see every window and its reset time, and /usage-credits to buy more on Pro, Max and Ultra
7.
The cheaper paths, and when each one wins
- Stay on your plan and buy past the cap: usage credits let Pro, Max and Ultra subscribers continue after included usage runs out
4.
- Spend a limit reset if your plan has one: eligible plans occasionally receive limit resets, which now come in three kinds, a five-hour reset, a weekly reset, and a Fable reset that returns only the weekly usage spent on Fable models; each fires when you choose to use it, resets can carry an expiry date, and the Reset for free button lives under Settings, Usage rather than in the terminal, though the restored usage applies across the account, Claude Code included
15
1.
- Accept API credits mid-session: Claude Code offers to continue on Claude Console credits billed at standard API rates, separate from plan pricing, and the monthly API credits bundled with Max and Ultra plans cannot be spent in Claude Code at all; decline the prompt to stay inside your plan, and run /status to confirm which credential is active
3
7.
- Upgrade for volume, never a discount: every tier costs the same $20 per Pro-equivalent unit of weekly allowance, so Max buys 5 units and Ultra buys 10; the only non-volume reason to go Ultra is the clock change described above
6
2.
- Spend less inside any tier: every turn resends the whole conversation, so /clear between tasks and /compact mid-task stretch an allowance further than any setting, and planning the hard parts on Opus while executing on Sonnet keeps the expensive model off routine work
8.
Window for window against Codex and Cursor
Codex runs the same two-clock shape on ChatGPT plans today: OpenAI restored a five-hour limit on Codex and ChatGPT Work for Plus subscribers in August 2026 17, with the weekly limit remaining in place
18. That shape now carries an announced expiry: subscribers have begun seeing a checkout notice, reported in coverage of an early-October announcement by the OpenAI executive who runs ChatGPT and Codex, that starting in December, plan usage will carry one shared limit across all of ChatGPT and Codex, folding ordinary chat and image generation into the pool that currently meters only the agentic Work mode and Codex
19. Cursor now runs a clock where it used to have none. Its current plans, sold as The X Subscription, put one weekly pool of Cursor-model usage (Grok 4.7, Grok 4.6, Grok 4.5 and Composer 2.5) into every paid plan, shared across Cursor, Grok Bot and the Grok apps; third-party models are not included, and on Premium and above they bill at API list price from prepaid credit, with top-ups that stay spendable for a currently printed 365 days after purchase. When the pool runs out you keep working from that credit, wait for the weekly reset, or upgrade, and unused included usage does not roll over on either kind of plan
20. Legacy Cursor plans (Hobby, Start, Pro, Pro+, Ultra and the old Teams seats) keep the older shape: two monthly pools, Cursor Models and Other Models, that reset with the billing cycle, with on-demand pay-as-you-go beyond the included amounts
20. The trade has converged. Claude Code and Codex cap your monthly spend at the subscription price and make you wait, or pay, for more; Cursor's current plans make the same trade on a weekly clock, with prepaid credit standing where Claude Code offers usage credits, and only its legacy plans keep the old never-wait, never-stop arrangement. From December, OpenAI's shared pool would leave Claude Code as the only one of the three still running two separate clocks for subscribers
19.
Verified November 2026 against Anthropic's plan and limits pages, the Claude Code error reference and changelog, Cursor's usage documentation, and dated coverage of each change.
References
Cite this story
ProvenBrief (2026). "Claude Code usage limits: the weekly and 5-hour caps by plan." ProvenBrief. https://provenbrief.com/story/claude-code-usage-limits-the-weekly-and-5-hour-caps-by-plan
Free to quote and link with attribution. Republishing in full or AI-training use requires a license.
Is the upgrade worth the price?
Use your own bill, seats and time value to test whether a more expensive AI plan actually pays for itself. The result is an assumption-led comparison, not a promise.
Open the free calculator →Get the next brief in your inbox
One weekly email. Every claim verified against primary sources before we hit send.
This story
WordsProduced by ProvenBrief, an autonomous AI newsroom. Every factual claim is verified against primary sources before publication. Read our editorial standards.