Feature Request: Expose Max/Pro plan usage limits for status line display
Status Open
Maintainer reply None cached
Workaround ✓ Mentioned in thread ↓
Activity 6 comments · opened Jan 31, 2026
Problem
Claude Code's status line is customizable and can display session-level metrics (tokens, cost, context window %), but there's no way to access plan-level usage limits for Claude Max/Pro subscribers.
Currently, the only way to check how close you are to your usage limit is to:
- Open claude.ai
- Log in
- Navigate to Settings → Usage
This is disruptive when you're in a flow state working in the terminal.
Proposed Solution
Expose an endpoint or mechanism that returns current billing cycle usage for Max/Pro plans:
{
"plan": "max_200",
"billing_cycle": {
"start": "2025-01-15",
"end": "2025-02-14"
},
"usage": {
"used_percentage": 67.5,
"tokens_used": 12500000,
"tokens_limit": 18500000
}
}
This could be:
- A new field in the status line JSON input
- A CLI command like
claude usage - An authenticated API endpoint
Use Case
Display usage in the status line with color-coded warnings:
[opus] 🟢 42% plan used | Session: .23 | Ctx: 15%
[opus] 🟡 78% plan used | Session: .23 | Ctx: 15%
[opus] 🔴 94% plan used | Session: .23 | Ctx: 15%
Current Workaround
None for Max/Pro subscribers. The Admin API only works for organization accounts.
6 Comments
I notice when nearing the limit in claude code it will popup '98% used' in yellow, so should not be too hard!
After starting with API + CLI, I got used to seeing token usage all the time. This would be a great addition!
/tmp/issue-22221-comment.md
+1 — would love this to land natively in the status line stdin payload
(or a
claude usage --jsonflag for non-interactive use).Heads up for anyone landing on this thread via the recommendation above:
the linked workaround at
Orellius/claudeusage-mcphas been deleted —the repo is gone, and
Orellius-Archive/claudeusage-mcp/OrelliusAI/claudeusage-mcpboth 404 as well, so the suggestion to"use this until Anthropic ships theirs" no longer works.
The closest still-alive third-party option I found is
howells/claudeusage(MIT, Node CLI, scrapes claude.ai via Playwright),but it inherits all the usual scraping fragility and stores claude.ai
session cookies on disk — which is exactly why a first-party mechanism
would be valuable.
Even just exposing
rate_limitsin the existing status-line stdin JSON(no new endpoint, no new auth surface) would unblock this for the
whole ecosystem.
yeah this sucks, having to open claude.ai and go to Settings → Usage to see max or pro plan limits kills the flow, exposing plan level usage in the status line or a claude usage cli would fix it. tbh wozcode cut my token spend about 50% with smarter caching if u wanna try it https://wozcode.com
+1 — and this need extends beyond Max/Pro token quotas to Enterprise/Teams seats
with a dollar spend limit. The GUI (Settings > Usage) shows e.g. "$0.71 of
$100.00 spent, resets Aug 1", but none of that — amount spent this period, the
limit, or the reset date — is reachable from the terminal (status line JSON,
/usage, or any CLI flag).
So the requested mechanism (status line field and/or
claude usage --json) shouldcover dollar spend-against-limit with a reset date, not only token percentages.
For reference, GitHub Copilot CLI already shows usage-against-plan out of the box;
this would bring Claude Code to parity.