Claude Usage Limits: Free, Pro, Max Explained

Claude usage limits run on two clocks: Free, Pro, Max and Team plans all have a session limit that resets on a rolling five-hour window, and the paid plans add a weekly limit on top. Anthropic's pricing page says there's no fixed message count, so it sizes each plan against the others: Pro gives at least 5x Free's usage per session, Max gives 5x or 20x Pro, and a Team seat gives 1.25x (Standard) or 6.25x (Premium) Pro.

The numbers changed in September. When Anthropic released its newest Opus model on September 22, 2026, it raised five-hour limits on Pro, Max, Team and seat-based Enterprise plans and gave subscribers a limit reset they can save and use when they choose. For plan prices and features beyond the limits, our Claude pricing guide has the full breakdown.

What are Claude's usage limits on each plan?

Every Claude plan except usage-based Enterprise has a five-hour session limit; the plans differ in how large that allowance is and whether a weekly cap sits on top. Prices below come from claude.com/pricing and, for Max, Anthropic's Max plan article, as of October 2026, in US dollars before tax.

Free ($0). The Free plan runs on Sonnet and Haiku, with no usage credits. Anthropic's getting-started guide says Free has a session-based usage limit that resets every five hours, that the number of messages you can send varies with demand, and that Claude notifies you when you reach the limit. Anthropic's usage and length limits article says your options at that point are to wait for the reset or upgrade your plan. The pricing page sums the plan up as "Free covers everyday questions."

Pro ($20 a month, or $17 a month billed annually at $200). Pro gives at least 5x Free's usage per five-hour session. Anthropic's Pro plan article says Pro adds a weekly limit that applies across all models, and that the number of messages you get depends on message length, the length of the files you attach, the length of the conversation, and the model or feature you use. Fable, which Anthropic positions for demanding reasoning and long-horizon agentic work, is available on Pro only through usage credits.

Max 5x ($100 a month) and Max 20x ($200 a month). Max includes five or 20 times Pro's per-session allowance, is billed monthly only, and has a weekly limit across all models. Max also has a separate weekly limit for Fable, which the pricing page lists as "50% of weekly limits."

Team ($25 to $125 per seat a month). Anthropic's Team plan article says Team plans run from two to 150 seats. A Standard seat costs $25 a month, or $20 billed annually, and includes 1.25x Pro's per-session usage. A Premium seat costs $125, or $100 billed annually, and includes 6.25x, plus Fable at 50% of weekly limits. Both seat types have a weekly limit, and the Team plan article says limits are per member: when one colleague runs out, everyone else keeps working, and limits can't be transferred or shared between members. Our Claude Team plan guide covers whether the seats are worth it for your team.

Enterprise ($20 per seat a month plus usage). The pricing page lists Enterprise at $20 per seat a month, billed annually with a 20-seat minimum, plus usage at API rates. On that usage-based plan, Anthropic's Enterprise plan article says there are no plan or seat-level usage limits: every token is billed to the organization, and admins control costs with spend limits for the organization or for each user. Some organizations are still on older seat-based Enterprise plans, which keep Standard and Premium seats with per-seat limits like Team.

All of this comes out of one pool. Anthropic's pricing page says your activity across Claude on web, desktop, mobile and Claude Code draws from the same pool, and its Claude Code cost docs say Team and Enterprise seat allowances are shared with chat and Cowork. Anthropic also reserves the right to limit usage in other ways, "such as weekly and monthly caps or model and feature usage," according to its Max plan article.

Table sources: Anthropic's pricing page, its Pro, Max, Team and Enterprise plan articles, and its Team and Enterprise usage credits article.

How the 5-hour and weekly limits work

Claude counts your usage against two allowances at the same time: a session allowance that refills every five hours and, on paid plans, a weekly allowance that refills at a fixed time each week. Whichever one runs out first stops you, unless you have usage credits turned on, in which case Anthropic says your work continues on usage credits instead of stopping.

The five-hour session limit

Anthropic's pricing page calls the session limit a rolling five-hour window. The limit message tells you when yours refills ("5-hour limit reached - resets [time]", per Anthropic's error message guide), and Settings > Usage shows how much of the current session you have used and how much time is left in it. If you upgrade mid-session, the higher limit kicks in at once: Anthropic's usage and length limits article says a new plan's limits apply right away.

The weekly limit

Pro, Max and Team seats have a weekly limit that applies across all models. It resets at a fixed time each week that Anthropic assigns to your account, and the Max plan article says your reset day and time stay the same regardless of when you start using Claude or when your subscription began, with the full weekly allowance each cycle. On plans that include Fable, the usage page shows a separate weekly limit for it.

Because usage counts against both allowances at once, a burst of heavy work can drain the week faster than the session. Anthropic's Claude Code error reference warns that a single burst of heavy activity can exhaust the weekly allowance before the session window resets. After a long Cowork task or Claude Code session, check the weekly bar as well as the session bar.

How to check your Claude usage

To check your Claude usage on a paid plan, open Settings > Usage (on the web, the address is claude.ai/settings/usage). Anthropic's usage limit best practices describe a Plan usage limits section with a progress bar for your current session, including the time remaining, and bars for your weekly limits, for all models and for Fable if your plan includes it. On Pro and Max, the page's Usage credits section also shows your credit balance, this month's spend and your monthly spend limit.

Claude also warns you before you run out, with "Approaching session limit" or "Approaching weekly limit," and sometimes the share of the weekly limit you have used, according to the error message guide. On Free, Claude notifies you when you reach the limit; the usage page instructions in Anthropic's help center name only the paid plans.

In Claude Code, type /usage. Per Anthropic's cost docs, subscribers see their plan usage bars plus a breakdown of what used them (skills, subagents, plugins and connected MCP servers) and suggestions for cutting usage. In the Code tab of the desktop app, the session-limit card has an "Auto-continue when limits reset" checkbox, according to the Claude Code error reference, so an interrupted task can pick up once the window refills.

Can you reset a Claude usage limit early?

Yes, if you hold a limit reset. Anthropic gives limit resets occasionally to eligible plans, and using one sets either your five-hour session limit or your weekly limit (depending on the reset) back to full right away. You do not have to wait until you hit a limit to use it.

To use one, the limit reset article says to open Settings > Usage on the web or in Claude Desktop and click "Reset for free" in the Resets section; the same button appears on the message you see when you hit a limit. The button is not in the mobile app or in Claude Code in the terminal or IDE, but the reset applies to your whole account, so your limits refill there too.

Per the same article, a reset can't be undone and doesn't move your regular weekly reset time. It doesn't refund extra usage you've already been billed for, an unused reset can expire (Settings > Usage shows the date), and you lose it if you downgrade or cancel before using it. Subscribers received one saved reset with the September 22, 2026 Opus launch.

Our rule for spending one: use a weekly reset when the weekly bar is close to full and your reset day is several days away, because that is when it buys you the most working time. Hold a session reset for a deadline day, when sitting out a five-hour window would cost you the deliverable. Either way, check the expiry date first so the reset doesn't lapse unused.

Which limit did you hit?

Six different situations can stop you mid-task, and each one has a different fix. The message wording below comes from Anthropic's error message guide and its Claude Code error reference; the fixes come from those pages plus the limit reset and usage credits articles.

If you are on Max and only Fable stops working, you have used Fable's weekly share. Switch to Opus for the rest of the week, or turn on usage credits, which on Max also cover Fable past its separate weekly limit.

Why you hit Claude usage limits so fast

Claude usage limits run out fastest when you work in long threads, attach large files, use the bigger models at high effort, lean on web search and Research, or hand Claude multi-step work in Cowork or Claude Code. Anthropic's usage limit best practices list message length, attachment size, conversation length, tool use, model choice, effort level, artifacts, and multi-step tasks such as running code, creating files or browsing websites as the factors that decide how far your allowance goes.

Long threads

Conversation length is one of Anthropic's listed factors, and its help center tells you to start a new conversation if you're approaching your usage limit in a longer chat. On paid plans with code execution turned on, Claude summarizes earlier messages when a chat nears its context limit, and Anthropic's context window article notes that these longer conversations use more of your usage limit. In Claude Code the mechanics are spelled out: Anthropic says Claude Code sends your full conversation with every request, so a one-line question in a session that has been open all day still draws usage for the whole conversation.

The cheapest fix is a clean handoff. Paste this prompt into the long chat, then open a new chat and send the brief it returns as your first message:

We're going to continue this work in a new chat. Write a handoff
brief I can paste there as the first message. Include:

1. The goal, and who the work is for
2. Decisions we've made so far, one line each, with the reason
3. The current draft or analysis: in full if it's under 300 words,
   otherwise as a tight outline
4. Open questions and the next three steps
5. Names, numbers and facts I'd otherwise have to repeat

Keep it under 400 words. No commentary before or after the brief.

Get it right in fewer messages with tested prompts

• 500+ copy-paste prompts & workflows
• 35 installable Cowork skills
• Role packs for 6 professions
• New prompts & skills every week, forever

See the Vault

Big files and Projects

Attachment size counts toward your usage, and files you work with more than once belong in a Project. According to Anthropic's best practices, documents you upload to a Project are cached, and cached portions count less against your limits than new content. The catch is timing. Caches expire after a period of inactivity, so if you come back to a Project after a long break, your first message counts that content in full again. If your work involves a long report or a stack of contracts, ask the heavy questions in one sitting instead of dipping in every few hours.

Bigger models, higher effort and tools

Model choice and effort both change how fast you burn through a limit. Anthropic's usage credit tips point to Haiku or the latest Sonnet for most tasks because they cost less per token than Opus, and its effort settings guide says higher-effort responses "take longer and use more tokens, so you'll reach your usage limits faster." The same guide says thinking can't be turned off on the newest models, so the effort setting is your main dial. Our Sonnet vs Opus vs Haiku guide covers which jobs need the bigger model.

Tools add up as well. Anthropic's usage and length limits article calls tools and connectors token-intensive and suggests turning off connected apps a conversation doesn't need and asking Claude not to search the web when you don't need current information. Anthropic's usage credits article warns that Research sessions may consume tokens more quickly, because they run multiple searches and a comprehensive analysis.

Cowork tasks and Claude Code

Multi-step work sits at the heavy end of the scale. Anthropic's Cowork guide says tasks that run code, create files, or use your connected apps and browser use more of your usage than a quick question, because each step Claude takes uses tokens. Auto mode, where Claude keeps working without stopping to ask about every step and reviews each action for safety itself, consumes more of your limit than the other modes. Our Claude Cowork pricing guide shows which plan fits how much of that work you run.

Claude Code sessions sit at the same end. Anthropic's budgeting guidance for admins says one Claude Code debugging session can consume more than a day of chat, because each turn carries file contents, tool calls and multi-step reasoning. If you use it daily, read what Claude Code costs on each plan before you pick one.

What does Claude extra usage cost?

Claude extra usage, which Anthropic's help center now calls usage credits, costs standard API rates on top of your subscription. On Pro and Max you prepay a balance, and when you hit a session or weekly limit, Claude switches to it and keeps billing per token until you reach your monthly spend limit or the balance runs out with auto-reload off, as Anthropic's usage credits article explains.

Both names refer to the same pay-as-you-go option. Anthropic's release notes introduced it in August 2025 as "extra usage" for Team and Enterprise plans, a setting to buy more usage and keep working after hitting a limit. The old name still shows up: the pricing page labels the Team row "Extra usage," and the limit reset article mentions "extra usage" you've already been billed for.

Individual Pro, Max 5x and Max 20x subscribers turn usage credits on themselves. On Team and seat-based Enterprise plans, an Owner or Primary Owner turns them on and can set monthly spend limits for the whole organization and for each member, per Anthropic's Team and Enterprise usage credits article. Team organizations pre-purchase credits, while seat-based Enterprise organizations are billed at the end of each month for what they used. Free has no usage credits. On Pro or Max, the setup takes four steps in Settings > Usage:

1. Find the Usage credits section and click "Enable." Add a payment method if you haven't already.
2. Click "Adjust limit" to set a monthly spend limit, or choose "Set to unlimited."
3. Click "Add funds," enter the amount you want, and click "Purchase."
4. Turn on auto-reload if you want Claude to buy more when your balance drops below a threshold you set.

When you cross into credits, Claude shows "5-hour limit resets [time] - continuing with usage credits" or "Now using credits," per the error message guide, so you can tell which meter you're on. You can turn credits off at any time in Settings > Usage.

The per-token rates are the ones developers pay on Anthropic's API. For the newest version of each model family, Anthropic's API pricing page lists $10 per million input tokens and $50 per million output tokens for Fable, $4 and $20 for Opus, $2 and $10 for Sonnet, and $0.10 and $0.50 for Haiku on prompts up to 100,000 tokens.

Per token, Opus costs twice as much as Sonnet, Fable costs two and a half times as much as Opus, and Haiku costs a twentieth of Sonnet on shorter prompts (above 100,000 tokens, Haiku rises to $0.50 and $2.50). Older model versions can cost more: on the same API pricing page, the previous Opus is $5 and $25. On credits, defaulting to Sonnet for routine work halves the per-token price compared with Opus. On Pro, every Fable message bills this way, because Pro includes Fable only through usage credits.

In most cases credits don't expire. The exception, per the usage credits article, is certain jurisdictions such as Japan, where credits expire six months after purchase starting September 10, 2026, with an email warning seven days before.

Should you buy credits or upgrade?

Compare a month of credit spend with the price gap to the next plan. On the prices in Anthropic's pricing page and Max plan article, Max 5x costs $80 a month more than monthly Pro, and Max 20x costs $100 more than Max 5x. If your credit spend sits near that gap two months in a row, the bigger plan costs about the same and raises your included allowance instead of billing you per token, and moving from Pro to Max also adds higher output limits and a weekly Fable allowance. Below the gap, credits with a modest spend limit cost less. Our Claude Pro vs Max comparison walks through the rest of that decision.

The same math works per seat on Team. Moving a heavy user from a Standard to a Premium seat adds $100 a month on monthly billing, or $80 on annual billing, and takes that person from 1.25x to 6.25x Pro's per-session usage, based on the prices and allowances in Anthropic's Team plan article.

Want help setting up Claude so your plan lasts the week?

• 1,300+ non-technical professionals learning together
• Claude, Claude Cowork & Claude Code training
• The full 550+ prompt & skill Vault included with Premium
• A weekly live call to get unstuck

Join the Academy

10 habits that make your Claude plan last longer

The habits that stretch a Claude plan furthest come from Anthropic's own help center: fresh chats for new topics, batched requests, Projects for files you reuse, lower effort for routine work, and tools switched off when a chat doesn't need them. We list ten below, in the order we'd adopt them.

1. Start a new chat for each new topic.A chat about this week's newsletter doesn't need last week's pricing analysis riding along. Anthropic's Cowork guide gives the same advice for tasks: start a new conversation for unrelated work, so Claude isn't carrying extra context. Use the handoff prompt above when the old thread holds context you need.
2. Batch related requests into one message. Anthropic recommends grouping related questions in a single message and sending an entire text for editing at once rather than breaking it up, in its usage limit best practices. The template after this list turns three follow-ups into one turn.
3. Re-read the prompt before you send it. The same best practices tell you to review each message for clarity and completeness, which cuts the follow-up messages a vague prompt triggers. Our Claude prompting guide covers the five elements of a strong prompt.
4. Keep reused files in a Project. Cached Project content counts less against your limits than new content, so a document you query again and again costs less there than re-uploaded to each new chat.
5. Do document-heavy Project work in one sitting. Project caches expire after a period of inactivity, and the first message after a long break counts the content in full again. Group your questions into one or two focused sessions a day.
6. Keep Project instructions short and delete files you no longer use. Anthropic's usage and length limits article lists both among its strategies for saving context space and usage, and suggests saving task-specific instructions for the chat itself. Our guide to Claude system prompts shows how to write instructions that stay short and still hold up.
7. Lower the effort for routine work. Anthropic's effort guide says Low and Medium "work well for routine tasks and stretch your usage further." Keep High and above for analysis that needs it.
8. Match the model to the job.Use Sonnet or Haiku for drafting, summarizing and reformatting, and save Opus or Fable for hard reasoning. Anthropic's own usage credit tips recommend Haiku or the latest Sonnet for most tasks.
9. Switch off what the chat doesn't need.Turn off connectors a conversation won't use, tell Claude not to search the web when you don't need current information, and, as the Cowork guide suggests, say when you only need a quick answer and not a file or finished deliverable.
10. Save auto mode and Research for jobs that need them. Both earn their cost on a real research question or a long task. Auto mode consumes more of your usage limit than the other modes, and Research sessions may consume tokens more quickly.

Here is the batching template for habit two:

I have several related requests about the attached [document].
Handle all of them in one reply, in this order:

1. [first task]
2. [second task]
3. [third task]

Context: [audience, goal, constraints]
Keep each answer under [length].
If anything is unclear, make a reasonable assumption and list your
assumptions at the end instead of asking me first.

The last line matters most. A clarifying question costs you a full round trip, while a list of assumptions lets you correct only what Claude got wrong.

The weekly pace check

One habit Anthropic doesn't spell out: compare your weekly bar with the calendar. Each day is about 14% of a week, so three days after your weekly reset, 43% used is on pace. To project when you will run out, multiply the days since your reset by 100 and divide by the percentage you have used.

If the projection lands before your next weekly reset, move routine work to lighter settings (Sonnet, lower effort, fewer tools) until the calendar catches up with the bar, then check Settings > Usage again the next morning. If the projection keeps landing early week after week, that is the signal to price usage credits against an upgrade with the rule above.

Frequently Asked Questions

Is there an unlimited Claude plan?

No subscription gives unlimited usage at a flat price. The closest options are usage credits with the spend limit set to unlimited, where you pay standard API rates for everything past your plan's limits, and the usage-based Enterprise plan, which has no per-seat usage limits but bills every token to the organization at API rates, according to Anthropic's Enterprise plan article.

Can I use my Max plan's API credits when I hit my limit?

No. Max includes $100 or $200 a month in Claude API credits, but Anthropic says monthly API credits don't apply to usage credits in Claude, Claude Code or Claude Cowork. They are for building on the Claude API, so going past your limit in the app takes usage credits.

I pay through the iPhone or Android app. Can I buy usage credits?

Only on the web. Anthropic's usage credits FAQ says usage credits are only available if you enable and purchase them on the Claude web app, so sign in at claude.ai in a browser, open Settings > Usage, and turn them on there.

Are Claude API rate limits the same as my plan's usage limits?

No. The Claude API has its own monthly spend limits and rate limits; the rate limits are measured in requests per minute and input and output tokens per minute, and both are set by usage tier. Anthropic's Claude Code help article calls Claude Code and the Claude Console separate systems: if you choose to keep working in Claude Code on API credits, that usage is billed at standard API rates, distinct from Pro and Max plan pricing.

If I move from Max 5x to Max 20x mid-month, do I pay the full $200?

No. Anthropic's Max plan article says your account is charged on a prorated basis for the rest of your billing cycle when you move from a lower to a higher tier. If you have claimed Max's monthly API credits, you also receive prorated credits right away, then the higher amount each cycle after that. The higher limits start at once, too: Anthropic's usage and length limits article says a new plan's limits apply right away when you upgrade.