Claude Usage Limits: How Plans and Resets Work
Claude’s paid plans have two limits that count at once: a session limit that resets every five hours and a weekly limit across all models. Claude chat, Desktop and Claude Code share them. Max 5x ($100 a month) and Max 20x ($200) give five or twenty times Pro’s per-session allowance.
Checked on 28 September 2026 against claude.com/pricing, Anthropic’s help center and the Claude Code and API docs. Sources are listed at the end.
TL;DR
- Plans: Pro $20/month ($17 billed yearly), Max 5x $100, Max 20x $200.
- Two clocks: a 5-hour session window and a weekly window. Your weekly reset is a fixed day and time set for your account.
- Shared: claude.ai, Claude Desktop and Claude Code all use the same limits.
- At the limit: wait, upgrade, or pay for usage credits at API rates.
- API: billed per token, with its own monthly spend caps per tier.
- Watch it: Settings › Usage,
/usagein Claude Code, or live in the menu bar with FavTray AI Usage, free.
What do the Claude plans cost, and how much usage do they include?
| Plan | Price (US) | Usage, as Anthropic describes it |
|---|---|---|
| Free | $0 | Basic access |
| Pro | $20/month, or $17/month billed yearly | ”More usage per session than the Free plan” |
| Max 5x | $100/month | Five times Pro’s per-session usage allowance |
| Max 20x | $200/month | Twenty times Pro’s per-session usage allowance |
| Team, Standard seat | $25/seat monthly, $20 yearly | More usage than Pro |
| Team, Premium seat | $125/seat monthly, $100 yearly | 5x a Standard seat |
Anthropic doesn’t publish a fixed number of messages or tokens for any plan. How far an allowance goes depends on the model, the length of the conversation and anything you attach. That’s why the useful number is the percentage of your own limit, not a message count.
Deciding between Pro and Max? Our Is Claude Max worth it? post works through the breakeven in hours of use and against the API.
How do the 5-hour and weekly limits work?
Every request counts against both windows at the same time.
- Session limit. Anthropic’s wording is that it “will reset every five hours”. Settings › Usage shows how much time is left in the current session.
- Weekly limit across all models. It “resets at a fixed time each week that is assigned to your account”, and that day and time stay the same no matter when you start using Claude or when your subscription began.
- Model limits. Claude Code’s error messages list separate Opus and Sonnet limits (“You’ve hit your Opus limit”), which apply only to that model family.
- One pool. Anthropic says your usage of claude.ai, Claude Code and Claude Desktop “counts towards the same usage limit”. A long chat in the afternoon leaves less for Claude Code in the evening.
When you run out, Claude Code tells you which limit it was and when it resets, for example “You’ve hit your session limit · resets 3:45pm” or “You’ve hit your weekly limit · resets Mon 12:00am”. Anthropic also gives eligible plans an occasional free limit reset, which you apply from Settings › Usage; it doesn’t change your weekly reset day.
What happens when you hit the Claude usage limit?
You have three options, according to Anthropic’s help center:
- Wait for the session or weekly reset.
- Upgrade, from Pro to Max 5x or Max 20x.
- Buy usage credits. Pro and Max users can switch them on under Settings › Usage and keep going after the included limit. Credits are billed at standard API rates, separately from the subscription, with a monthly spending cap you set. In Claude Code,
/usage-creditsopens the same settings.
Switching models only helps for a model limit. The session and weekly limits cover every model. If you hit your Opus limit, /model to Sonnet keeps you working; if you hit the weekly limit, it won’t.
Watch the login. If Claude Code is signed in with a Console (API) account rather than your Pro or Max login, usage is billed to the API instead of your plan. /status shows which one you’re using.
How much does the Claude API cost?
The API has no 5-hour or weekly window. You pay per token, and your organisation’s tier sets rate limits and a monthly spend cap.
| Model | Input, per 1M tokens | Output, per 1M tokens | Cache hit |
|---|---|---|---|
| Haiku 4.5 | $1 | $5 | $0.10 |
| Sonnet 5 | $2 | $10 | $0.20 |
| Opus 5.5 | $4 | $20 | $0.20 |
| Fable 5.1 | $10 | $50 | $0.25 |
The Batch API is 50% off. Monthly spend caps are $500 on the Start tier, $1,000 on Build and $200,000 on Scale, and you can set your own lower limit in the Console under Settings › Billing › Spend limits. When a cap is reached, requests fail with HTTP 429 until the first day of the next month (UTC).
For Claude Code on the API, Anthropic’s cost docs give an average of “around $13 per developer per active day” across enterprise deployments, below $30 a day for 90% of users. Your number depends on model and context size; reduce Claude Code token usage covers the levers, and tracking Claude API costs shows where to read the bill.
How do you check your Claude usage?
- claude.ai: Settings › Usage shows bars for the current session and the week, with time remaining.
- Claude Code:
/usageshows the same plan bars plus activity for the last 24 hours or 7 days. It also warns you near the limit, for example “You’ve used 85% of your session limit”. - API: the Console’s usage and cost pages.
The catch with all three is that you have to go and look. Our guide to checking Claude Code usage limits covers the status line option too.
How do you track the Claude usage limit from the Mac menu bar?
FavTray’s AI Usage tracker puts the limits where you can see them while you work. It’s free on every install, with no daily limit and full history. For Claude it shows:
- Both windows: the 5-hour session and the weekly limit, plus any Opus or Sonnet weekly limit your plan reports, and usage credits if you have them.
- Pace: a tick on each bar where an even pace would put you, and a plain line: “Using faster than average”, “On pace” or “Plenty of room”.
- Time to limit: once you’re 10 minutes into a window, a projection such as “At this pace you’ll hit the limit in 1h 20m, around 16:40”, or the percentage you’ll finish the window on.
- Alerts at 80%, 95% and when a limit is hit.
- In the menu bar itself, optionally: the session or weekly percentage left, or the time to reset.
It reuses the Claude Code login already on your Mac, read-only. macOS may ask you once to allow access to the “Claude Code-credentials” keychain item. FavTray never refreshes or rewrites that login, so it can’t sign you out of Claude Code. The same screen tracks Codex, Cursor, Gemini, Copilot and API accounts if you use more than one tool, and the developer cockpit moves an AI limit at risk to the top of its “Needs you now” list. Want the numbers in the notch instead? FavShelf shows Claude and Codex usage bars under the camera, also free.
FavTray’s other tools have free daily limits; a one-time Lifetime licence removes them. AI Usage doesn’t need it.
Sources, all checked 28 September 2026: Claude pricing; Anthropic help center: What is the Max plan?, What is the Pro plan?, Using Claude Code with your Pro or Max plan, usage limits, usage credits, checking usage, limit resets; Claude Code docs: errors and costs; Claude API docs: pricing and rate limits.
Frequently Asked Questions
How do Claude usage limits work?
Paid Claude plans have two limits that count at the same time: a session limit that resets every five hours, and a weekly limit across all models that resets at a fixed day and time assigned to your account. Some plans also have separate weekly limits for Opus or Sonnet. Claude chat, Claude Desktop and Claude Code all draw from the same limits.
How much more usage does Claude Max give than Pro?
Anthropic sells Max in two sizes: Max 5x at $100 a month and Max 20x at $200 a month, with five or twenty times Pro's per-session usage allowance. Pro is $20 a month in the US, or $17 a month billed yearly. Anthropic doesn't publish a fixed number of messages or tokens for any plan, because usage depends on the model, the length of the conversation and attached files. Prices checked 28 September 2026.
When does the Claude weekly limit reset?
At a fixed time each week that Anthropic assigns to your account. It doesn't move when you start using Claude or when your subscription renews. Settings › Usage on claude.ai and /usage in Claude Code both show when your limits reset.
What happens when I hit the Claude usage limit?
You can wait for the limit to reset, upgrade your plan, or turn on usage credits in Settings › Usage, which are billed at standard API rates on top of your subscription. In Claude Code, a session or weekly limit applies to every model, so switching models doesn't help; an Opus or Sonnet limit only applies to that model family, so /model can keep you working.
How much does the Claude API cost?
API usage is billed per million tokens. On 28 September 2026 Anthropic's pricing page listed Haiku 4.5 at $1 input and $5 output, Sonnet 5 at $2 and $10, Opus 5.5 at $4 and $20, and Fable 5.1 at $10 and $50. The Batch API is 50% off, and cache hits cost far less than fresh input.
How can I see my Claude usage limit on a Mac without opening a browser?
FavTray's free AI Usage tracker shows your 5-hour session and weekly limits in the Mac menu bar, with a pace line, a projected time to limit and alerts at 80%, 95% and when a limit is hit. It reuses the Claude Code login already on your Mac, read-only.