Claude Usage Limits Explained: 5-Hour and Weekly Limits

Claude’s paid plans have two usage limits: a five-hour session limit and a weekly limit. Both are shared by claude.ai, the desktop and mobile apps, and Claude Code. Anthropic does not publish either limit as a number of tokens or hours. It publishes rules and multipliers: Max 5x gives five times Pro’s per-session usage, and Max 20x gives twenty times. When you hit a limit, you can wait for the reset, buy past it with usage credits at API rates, or upgrade.

Every rule below was checked on October 25, 2026 against the Claude Help Center, claude.com/pricing and the Claude Code docs. Each row links to its source. Claudekit.ai is an independent community site, not affiliated with Anthropic.

Claude limits at a glance

• One allowance across Claude and Claude Code.Usage on claude.ai, Claude Desktop, the mobile apps, Claude Code in the terminal and Claude Code in IDEs counts toward the same limits. (support.claude.com)
• Five-hour session limit.Paid plans have a session limit that resets every five hours. (support.claude.com)
• Weekly limit with a fixed reset time.Paid plans also have weekly limits that reset once a week. The reset day and time are assigned to your account and stay the same regardless of when you start using Claude or when your subscription began. (support.claude.com)
• What uses the allowance.Message and file length, conversation length, features used, model choice and effort level all change how fast you use your allowance. Cached content (for example project knowledge you reference again) counts less than new content. (support.claude.com)
• Other caps at Anthropic's discretion.The Pro plan article says Anthropic may also limit usage in other ways, such as weekly and monthly caps or model and feature usage, to manage capacity. (support.claude.com)
• Wrap-up allowance in Claude Code.If you reach the five-hour limit mid-response, Claude Code may keep working briefly to finish the step in progress. On Pro this applies up to once per weekly usage period; on Max it can apply each time you reach a five-hour limit, within your weekly limit. (support.claude.com)
• Session and weekly limits apply across models."You've hit your session limit" and "You've hit your weekly limit" apply across all models, so switching with /model does not help. After a model-specific message such as "You've hit your Opus limit", switching to a model outside that family does. (code.claude.com)
• Usage credits bill past the limit at API rates.Pro, Max 5x and Max 20x subscribers can turn on usage credits to keep working after reaching the included limits. Usage credits are billed at standard API rates, with an optional monthly spend limit. (support.claude.com)
• An API key in your environment bills API rates.If an ANTHROPIC_API_KEY environment variable is set, Claude Code uses it instead of your subscription, and you pay API usage charges rather than using the plan's included usage. (support.claude.com)
• Max and Team API credits do not cover Claude Code.The monthly API credits on Max and Team cover the Claude API, Managed Agents, the Agent SDK and the Console playground, but not Claude Code or extra usage in the Claude apps. They do not change your usage limits. (platform.claude.com)
• No published token or hour numbers.Anthropic's usage-limit articles describe the five-hour and weekly windows and the multipliers between plans, but do not state either limit as a number of tokens or hours. (support.claude.com)

Limits by plan

Every row was checked against its source on Oct 25, 2026. Anthropic changes limits often; if a source no longer matches, the official page wins.

The pattern: Free has the smallest allowance and no Claude Code. Pro adds Claude Code and the two windows. Max multiplies Pro’s session allowance and adds a separate weekly limit for Fable models. Team seats have their own allowance per member, with Premium seats at five times Standard. Usage-based Enterprise has no per-seat limits and bills usage at API rates instead.

For prices and features beyond limits, see our Claude pricing guide and Claude Pro vs Max.

Claude Pro usage limits

Pro costs $20 a month billed monthly, or $17 a month billed annually ($200 up front). Its usage limits work like this:

• Session limit: a five-hour window, shared by Claude chat and Claude Code. Pro gives more usage per session than Free.
• Weekly limit: one weekly limit across all models, resetting at a fixed day and time assigned to your account.
• No published number. Anthropic doesn’t state Pro’s allowance in messages, tokens or hours, so any figure you see elsewhere is someone’s measurement, not a rule.
• Wrap-up allowance: if you reach the five-hour limit mid-response in Claude Code, it may keep working briefly to finish the step. On Pro this applies up to once per weekly usage period; on Max it can apply each time.
• Past the limit: wait for the reset, or turn on usage credits billed at standard API rates, with an optional monthly spend limit.

Max is defined relative to Pro: Max 5x gives five times Pro’s per-session usage and Max 20x gives twenty times, and both add a separate weekly limit for Fable models. If you hit Pro’s limits every week, Claude Pro vs Max walks through whether the upgrade is worth it.

What resets, and when

• The session limit resets every five hours. The Claude Code status line docs describe it as a rolling window, and Claude Code shows the exact reset time when you hit it.
• The weekly limit resets once a week at a day and time assigned to your account. It does not move when you start working, and you get the full weekly allowance each cycle.
• Model-specific limits, such as Max’s weekly Fable limit, sit on top of the two shared windows.

Anthropic has changed the size of these windows several times. The Claude Code weekly limit, for example, became 25% higher than its pre-promotion level on September 14, 2026. The full list, with sources, is in the Claude limits changelog.

What counts against your limit

Usage is not counted in messages. The Help Center lists what makes it go faster:

• Length: long messages, big attachments and long conversations.
• Model and effort: larger models and higher effort levels use more.
• Features: tools, code execution and agentic work send more requests.

Cached content counts less than new content. In Claude chat that means reusing Project knowledge instead of re-uploading files.

Claude Code adds a few patterns of its own, from Anthropic’s costs documentation:

• Every request re-sends the whole conversation, so a one-line question in a session that has been open all day still draws on the full context.
• Your first message after a break longer than the cache lifetime misses the cache and re-processes everything.
• Subagents, agent teammates, scheduled /loop tasks and idle goal check-ins each send their own requests.
• /compact is itself a large request. /clear costs nothing.

Our guide on how to reduce Claude token usage turns these into habits.

Claude Code limits

On a Pro, Max, Team or seat-based Enterprise plan, Claude Code has no separate allowance: it draws on the same five-hour and weekly limits as Claude chat. What differs is how fast it uses them, because agentic work re-sends your context and calls tools on every step (see the list above). Three details apply only to Claude Code:

• Weekly limits in Claude Code are higher than they were. From September 14, 2026 they are permanently 25% above their level before the May 2026 promotion. Five-hour limits were not part of that change. The limits changelog has the history.
• Claude Code names the limit you hit, so you know whether switching models helps. The messages are in the table under “What to do when you hit a limit” below.
• On an API key there are no plan limits. If Claude Code signs in with a Claude Console key, or ANTHROPIC_API_KEY is set, you pay per token instead. Your ceiling is then your organization’s API rate limits and any workspace spend limit you set in the Console. The Claude Code cost docs suggest per-user token and request limits by team size.

To watch Claude Code limits as you work, put the five-hour and weekly percentages in your status line. To make them last longer, see how to reduce Claude token usage.

How to check your usage

Where	What you see
claude.ai, Settings > Usage	Progress bars for the five-hour session and the weekly limit, reset times, and usage credits if you use them
Claude Code, /usage	Plan usage bars, a breakdown of what used them (skills, subagents, plugins, MCP servers) and recommendations. /cost and /stats open the same screen
Claude Code status line	rate_limits.five_hour and rate_limits.seven_day percentages and reset times, always on screen (claude.ai Pro and Max only). See the statusline recipe

The /usage breakdown is computed from session history on that machine, so it does not include usage on claude.ai or other devices. The bars come from your account and do. The Help Center article on Claude Code with Pro and Max also mentions /status; in current Claude Code, /status shows your version, model and account, and /usage shows the limits.

What to do when you hit a limit

Claude Code tells you which limit you hit:

Message	Meaning	What helps
“You’ve hit your session limit”	Five-hour window, all models	Wait for the reset, or use usage credits
“You’ve hit your weekly limit”	Weekly window, all models	Wait for the reset, use usage credits, or upgrade
“You’ve hit your Opus limit” (or another family)	A model-specific limit	Switch with /model to a model outside that family

Your options, cheapest first:

1. Wait. Run /rate-limit-options. On Claude Code v2.1.234 or later it can wait and continue the interrupted task after the reset.
2. Turn on usage credits. Run /usage-credits. On Pro and Max it opens Settings > Usage, where you can set a monthly spend limit. Credits are billed at standard API rates.
3. Upgrade. Pro to Max 5x, or Max 5x to Max 20x. Do it after you hit limits repeatedly, not after one bad day. The plan picker compares your usage, priced at API rates, with each plan.
4. Move automation to the API. CI jobs and scripts on a Console key do not touch your plan allowance.

Two traps that cost money

• An API key in your environment. If ANTHROPIC_API_KEY is set, Claude Code uses it instead of your subscription and bills API rates. Unset it and run /login. Our Claude Code pricing guide shows how to check.
• Expecting Max’s API credits to cover Claude Code. The $100 or $200 a month in API credits on Max covers the Claude API, Managed Agents, the Agent SDK and the Console playground. It does not cover Claude Code or extra usage in the apps, and it does not raise your limits.

Anthropic also says it may limit usage in other ways, such as monthly caps or model and feature limits, to manage capacity. Treat the published rules as the floor of what can change, not a contract.