Claude Code Usage: How to Check It and What Each Plan Allows

#Two limits, one shared pool

Anthropic's wording is short: "Both Pro and Max plans have a five-hour session limit and a weekly limit." The session limit "will reset every five hours". The weekly limit "applies across all models" and "resets at a fixed time each week that is assigned to your account", so it does not depend on when you subscribed or when you started working. Max plans add one more: "a separate weekly limit for Fable."

The detail that surprises people is the sharing. These limits "are shared across Claude and Claude Code", and use inside an IDE counts too. An afternoon of long conversations on claude.ai leaves less for the terminal, and the reverse.

The limits also apply across models, so switching with /model does not bring access back after a session or weekly limit. The exception is a model-specific message such as "You've hit your Opus limit": switching to a model outside that family lets the work continue.

#What each plan allows

From Anthropic's help center, read 21 October 2026. The number of messages "will vary based on message length", attachments, conversation length and the model used, which is why Anthropic gives multiples instead of counts.

#How to check Claude Code usage

/usage inside Claude Code. For Pro, Max, Team and Enterprise subscribers it shows plan usage bars, activity stats and a usage breakdown. The breakdown attributes usage to skills, subagents, plugins and MCP servers, recommends changes for anything behind at least 10 percent of it, and switches between the last 24 hours and 7 days with d and w. That breakdown is "computed from local session history on this machine", so use on other computers or on claude.ai is not in it. The dollar figure in its Session block is an estimate at list price meant for API users; for subscribers it "isn't relevant for billing purposes".

/status. Anthropic's plan guide also says to "Monitor your remaining allocation using the /status command."

Settings > Usage on claude.ai. It shows your next reset time and, if you have turned them on, usage credits. Team and Enterprise admins see member usage under Organization settings > Usage.

If you use an API key. With ANTHROPIC_API_KEY set in your environment, Claude Code uses the key instead of your subscription and you pay API rates. Actual charges for that are on the Claude Console usage page, not in the plan meters.

#Put the numbers in the status line

Claude Code's status line is a bar at the bottom of the session that "runs any shell script you configure", receiving session data as JSON on stdin. It runs locally and uses no tokens. The quickest setup is to describe what you want: /statusline show model name and context percentage with a progress bar writes a script into ~/.claude/ and updates your settings. By hand, add a statusLine entry to ~/.claude/settings.json:

{ "statusLine": { "type": "command", "command": "~/.claude/statusline.sh", "padding": 2 } }

For Pro and Max subscribers the JSON includes rate_limits.five_hour.used_percentage and rate_limits.seven_day.used_percentage, each with a resets_at time, which appear after the first response in a session. A script that prints both windows can be this short, using jq:

#!/bin/sh
# ~/.claude/statusline.sh: reads Claude Code's session JSON on stdin
jq -r '"5h \(.rate_limits.five_hour.used_percentage // "?")%  week \(.rate_limits.seven_day.used_percentage // "?")%"'

The same JSON carries cost.total_cost_usd, an estimate computed at list price, and the context window percentage.

#When you hit the limit

Claude Code warns you as you approach any limit. If you reach one in the middle of a step it may finish that step and show "Usage limit reached · wrapping up". Anthropic calls that allowance "small and capped": on Pro it applies at most once per weekly period, on Max at each five-hour limit within the weekly one, and it is still rolling out.

After that there are four ways forward: wait for the reset, upgrade (Pro to Max 5x, Max 5x to Max 20x), turn on usage credits, or switch to a Claude Console account with API credits. Usage credits move you to "consumption-based pricing at standard API rates" until your monthly spend limit or balance runs out, and they do not change when your limits reset. Claude Code 2.1.234 and later can also wait for the reset and pick the interrupted task back up; the choice lives in /rate-limit-options.

One trap worth knowing: the monthly API credits that come with Max plans "can't be used in Claude Code or the Claude apps".

#Trackers outside Claude Code

Several tools read the session logs Claude Code keeps on disk and turn them into running totals. ccusage, an open-source command-line tool, reports tokens and estimated cost by day, month or session from those logs, and menu bar apps do the same with a meter in view. Working from local logs, they see one machine and they estimate; the /usage bars and the Settings page are the authority on your plan. Hooks that post a notification when Claude Code needs you are covered in how to monitor Claude Code on a Mac.

NotchBay, the app this site is about, takes the same approach on a notched MacBook: the notch shows the running agent's status and the 5-hour and weekly usage, computed from the local Claude Code logs, and permission prompts can be approved or denied from the notch. How Claude Code's limits compare with Codex and Cursor is laid out in Claude Code vs Codex vs Cursor.

#Frequently asked questions

#What this means for the notch

The weekly limit resets at a time Anthropic assigns and the session limit on its own five-hour clock, so neither lines up with a working day. Most people learn where they stand from a warning, the least useful moment to find out. A number that stays on screen, in the strip beside the camera that the menu bar leaves empty, makes the window something to schedule around: start the big refactor when the bar is low, not when it is nearly full.