Claude Code usage limits
Claude Code doesn't have limits of its own on Pro and Max. It draws from your plan's session and weekly limits, the same pool as your chats.
The short answer
On Pro and Max, Claude Code shares your plan's 5-hour session limit and weekly limit with claude.ai and Claude Desktop, including Claude Code inside VS Code, Cursor, and JetBrains. Run /usage to see your plan usage bars. At the limit you can wait, turn on usage credits, upgrade, or switch to API billing through the Claude Console.
Claude Code limits at a glance
| Plans | Pro, Max 5x, Max 20x (not Free) |
|---|---|
| Limits | Your plan's 5-hour session limit and weekly limit; Max also has a weekly Fable limit |
| Shared with | claude.ai, Claude Desktop, and Claude Code in VS Code, Cursor, and JetBrains |
| Check it | /usage shows plan usage bars and what's using them |
| Mid-response at the limit | A small wrap-up allowance lets the current step finish |
| At the limit | Wait, usage credits (/usage-credits), upgrade, or Console API billing |
| Watch out | ANTHROPIC_API_KEY in your shell switches Claude Code to API billing |
Seeing what's left from inside Claude Code
/usage shows your plan usage bars, plus a breakdown of what's eating them: skills, subagents, plugins, MCP servers, and scheduled loops, for the last 24 hours (d) or 7 days (w). The breakdown only covers sessions on this Mac, not claude.ai.
The status line gets your session and weekly percent and reset times after every reply, so a script can show them under the prompt all the time. Our status line guide shows how.
Why a long session burns through it
Claude Code sends the whole conversation with every request, and again with every batch of tool results. A one-line question in a session that's been open all day still pays for the whole history. /clear between unrelated tasks is the biggest saver. Subagents, agent teams, scheduled loops, and /compact on a huge context all add up too.
What happens when you hit it
"You've hit your session limit" and "You've hit your weekly limit" cover every model, so /model won't get you around them. The message tells you when the window resets. A model-specific message, like "You've hit your Opus limit", is different: switching to a model outside that family keeps you going.
Your options: wait (Claude Code 2.1.234 and later can wait for the reset and continue the interrupted task on its own), turn on usage credits with /usage-credits, upgrade (Pro to Max 5x, or Max 5x to 20x), or use API credits through a Claude Console account, billed at API rates.
Will you make it to the reset?
A percentage tells you how much you've used. It doesn't tell you whether, at the pace you're going, you'll run out before the window resets. Fuel Gauge does that math: a free Mac menu bar app that reads your Claude Code limits after every reply and says "Makes it to reset" or "Runs out around 3:40p, 1h 20m early". It can put the same line under your Claude Code prompt.
Sources
Checked on October 20, 2026. Anthropic changes plans from time to time, so the help pages are the final word.
Questions
Does Claude Code have its own usage limit?
Not on Pro and Max. It shares your plan's session and weekly limits with claude.ai and Claude Desktop.
How do I check my Claude Code usage?
Run /usage in Claude Code for your plan usage bars and a breakdown of what used them, or see Settings › Usage on claude.ai.
Is Claude Code included in the Free plan?
No. Claude Code comes with Pro and Max.
Why did Claude Code bill my API key?
If ANTHROPIC_API_KEY is set in your environment, Claude Code uses it instead of your subscription. Unset it and log in with your Pro or Max account.
Keep reading
Know if you'll make it to reset.
Free. No account, no subscription.
brew install --cask coreyhaines31/tap/fuelgaugeOr ask your coding agent:
Install Fuel Gauge on my Mac: run `brew install --cask coreyhaines31/tap/fuelgauge`, open Fuel Gauge, then run `fuelgauge install-statusline` and `fuelgauge status`.
macOS 14 Sonoma or later · Apple Silicon and Intel