How much Claude Code quota is left?
In Claude Code, /usage shows how much of each plan limit you have used and when it resets. Agent HUD keeps the same windows at the edge of your screen while you work in other apps, and tells you whether each one lasts until its reset.
Updated 2026-10-26
Where Claude Code shows it
/usage shows your plan's usage bars and activity stats. On Pro, Max, Team and Enterprise plans it adds a breakdown of what counts against your limits, by skill, subagent, plugin and MCP server, with suggested changes. /cost and /stats open the same screen, and it opens while Claude is still responding.
The cost at the top of /usage is an estimate at list price, meant for API users. On a subscription it does not change what you pay. On claude.ai, the Usage page in Settings is where usage credits are turned on and their spend is checked.
In the status line
A status-line script receives rate_limits.five_hour and rate_limits.seven_day, each with used_percentage from 0 to 100 and resets_at in Unix seconds. They appear for claude.ai Pro and Max subscribers after the session's first response. Either window can be missing, and a window drops out once its reset time passes, so a missing value means unknown.
Behind a Claude apps gateway with a spend limit, the same object carries a spend_limit window. A status line covers the one terminal it runs in.
Which limits apply?
A subscription is metered in a five-hour session window and a weekly window, both shared across every model. A plan can also have a limit for one model family, such as Opus or Sonnet. Each window has its own reset time, and any of them can be the one that stops you.
On Team and Enterprise plans the allowance belongs to each seat, and Claude chat and Cowork draw on it too.
What Agent HUD shows
Each window is a row: the share used, or the share left if you prefer, and the time it resets. The glow is green under 70% used, yellow from 70% and red from 90%. Hovering shows every window with its reset time or burn rate, and in Simple mode Claude's ring follows its most used window. A model family's weekly window appears when your plan reports one.
Agent HUD asks the Claude Code already on your Mac for the signed-in account's usage, with hooks off. No message is sent to a model.
An API key or a third-party platform has no plan limits to read, and the row says so. If Claude Code is signed out, the row says that instead. Tokens and sessions are still counted from the records on your Mac.
Will it last until the reset?
The burn rate is the share of a window used per hour in its current cycle. If that pace empties the window before its reset, the row shows when it runs out; otherwise it shows the share you will have used by the reset.
Agent HUD alerts you once when a window is at risk, because the forecast says it runs out before its reset or because it crosses 90%. It alerts you again when the window is used up, and when a new reading confirms the reset.
Which session used the most?
Agent HUD's statistics count each Claude Code session's tokens, its sub-agents included, with its turns, calls, context fill and cache hits, and an estimate at Anthropic's list price marked ≈. The estimate is not a charge.
When a limit is reached
Claude Code's /rate-limit-options lists the ways to keep working: wait and continue the interrupted task once the limit resets, add usage credits, or move to a larger plan.
After a model-family limit, a model outside that family keeps working; the session and weekly limits cover every model. A context warning is not a usage limit: it means the conversation is close to the point where Claude Code summarizes older history.
Agent HUD shows when each window resets, and tells you when a new reading confirms the reset.
Questions
Is the percentage used or remaining?
Claude Code shows the share used. Agent HUD shows the share used by default and can show the share left instead; its colours and alerts always follow the share used.
Why does Agent HUD show no quota for my Claude Code?
An API key or a third-party platform has no plan limits, and a signed-out Claude Code has none to report. The row names which one it is, and tokens and sessions are still counted.
Why does a window still show after its reset time?
Agent HUD keeps the last reading until a new one confirms the reset, and says it is waiting for an update. A clock passing the reset time does not prove the allowance came back.
Does reading the quota send anything to Claude?
Agent HUD asks the Claude Code on your Mac for the account's usage, with hooks off. No message goes to a model.
Can I see it on my phone?
Yes. On the same network as your Mac, iPhone and Apple Watch show your quotas without a subscription. Agent HUD Remote shows them on any network.