Claude Code /usage Command Explained: Know Your Remaining Quota

Why You Need /usage

What’s the worst thing that can happen when coding with Claude Code?

It’s not an unfixable bug or an impossible feature request — it’s getting rate-limited right in the middle of your work.

Claude Code’s Pro and Max subscriptions come with usage limits, and all Claude products (Claude.ai, Claude Code, Claude Desktop) share the same quota pool. Spend half the day chatting on the web interface, and you might find your terminal quota nearly drained.

How do you know how much quota you have left? That’s exactly what /usage is for.

What Is /usage

/usage is a built-in slash command in Claude Code that shows your current usage and quota status.

In interactive mode, simply type:

/usage

It displays your consumption within the current rate-limit window, including:

Used quota: How much you’ve consumed in the current window
Remaining quota: How much you have left
Reset time: When the rate-limit window refreshes

In short, /usage answers one question: How much can I still use?

Understanding the Quota System

Before diving into how to use /usage, let’s understand how Claude Code’s quota system works.

Quota by Plan

Plan	Quota Multiplier	Approx. Tokens (per 5 hours)
Pro	1x	~44,000 tokens
Max 5x	5x	~88,000 tokens
Max 20x	20x	~220,000 tokens

Shared Quota Pool

Key point: All Claude products share the same quota. Your consumption on the Claude.ai web interface and in the Claude Code terminal draw from the same account.

That’s why sometimes you think “I barely used Claude Code, how am I already at the limit?” — it’s probably because you spent a bunch of tokens asking questions on the web earlier.

Tokens Add Up Fast

A seemingly simple operation like “edit this file for me” can actually consume 50,000 to 150,000 tokens. Every interaction in Claude Code sends the system prompt, conversation history, file contents, and tool call results — all of which count as tokens.

On top of that, a single Claude Code instruction can trigger multiple API calls: reading files, editing code, running tests, fixing bugs, running tests again — a simple “fix lint errors” command might generate 8 to 12 API calls behind the scenes.

Claude Code has several usage-related commands — don’t mix them up:

Command	Purpose	Best For
`/usage`	Check quota progress and remaining allocation	Pro / Max subscribers
`/cost`	View token consumption and cost for current session	API pay-as-you-go users
`/stats`	View usage patterns and statistics	All users
`/context`	Check current context window utilization	All users

/usage: How much quota do I have left? (quota dimension)
/cost: How much did this session cost? (cost dimension, API users only)
/stats: What do my usage patterns look like? (statistics dimension)
/context: How much room is left in the current context window? (context dimension)

Practical Scenarios

Scenario 1: Check Before a Long Session

Before starting a big task, run /usage to make sure you have enough quota. Nothing’s more frustrating than getting rate-limited halfway through.

/usage

If you’re running low, consider:

Waiting for the rate-limit window to reset before starting
Switching to the Sonnet model to save quota: /model sonnet

Scenario 2: Rate-Limited? Check When It Resets

Hit a rate limit unexpectedly? Run /usage to see the reset time.

The rate-limit window is typically 5 hours. Once it resets, your quota restores automatically. In the meantime, you can:

Work on tasks that don’t require Claude Code
Review your recent conversations for optimization opportunities

Scenario 3: Align Quota Across Products

You’ve been chatting extensively on Claude.ai and are about to switch to Claude Code for some real work. Check first:

/usage

See how much is left. Don’t burn through your quota on the web and then discover you can’t write a single line of code in the terminal.

Tips for Saving Quota

Tip 1: Compress Context with /compact

As conversations grow longer, each interaction consumes more tokens. Use /compact to compress the context — it significantly reduces token consumption for subsequent messages.

/compact

Tip 2: Use /clear Promptly

After completing a task, use /clear to reset the context before starting a new one. Don’t pile everything into a single conversation — the longer the context, the higher the cost per interaction.

Tip 3: Switch Models

Not every task needs Opus. For simple code edits and formatting tweaks, Sonnet is more than enough — lower consumption, faster responses.

/model sonnet

Tip 4: Make Good Use of CLAUDE.md

Document your project’s architecture, coding standards, and common commands in CLAUDE.md. This way you don’t have to describe the background in every conversation — saving those tokens for what actually matters.

Advanced: Real-Time Usage in the Status Bar

Running /usage manually every time is tedious. You can display usage information in real time right in the status bar.

Configure the status option in Claude Code’s settings to show quota information at the bottom of your terminal. This way you can always see your remaining quota without interrupting your workflow.

Final Thoughts

/usage solves a simple problem: keeping you informed about your quota.

When developing with Claude Code, tokens are your fuel. Hitting the road without knowing how much gas you have left means you’ll eventually break down mid-journey.

Build the habit — check /usage before starting work, check /usage during long sessions, check /usage when you get rate-limited. Three words: stay informed.

One command to avoid every “suddenly rate-limited” surprise.