Claude how to check token usage: app, Claude Code and API

Claude how to check token usage: it depends on the surface you work in. In the Claude app, open Settings > Usage. In Claude Code, run /usage. On the API, open the Usage page in the Claude Console. As of October 26, 2026, these are the official places Anthropic's docs point to, and each one answers a slightly different question.

This guide is for anyone who has hit a limit, or wants to avoid hitting one. It covers each place to look, what the numbers mean, and one thing that trips people up: a plan limit and a token bill are not the same number.

Check usage in the Claude app

Per Anthropic's usage limit best practices article, go to Settings > Usage, or open claude.ai/settings/usage. It applies to Pro, Max, Team and seat-based Enterprise plans.

At the top you see plan usage bars:

• Current session: how much of your five-hour session limit you have used, and how long until that session resets.
• Weekly limits: when the weekly limit resets, shown for all models.
• Usage credits: if you use them, how much you have spent.

The page shows bars and reset times, not token counts. As of October 26, 2026, Anthropic's usage and length limits article says paid plans have a session limit that resets every five hours, plus weekly limits. Usage across claude.ai, Claude Code and Claude Desktop counts toward the same limit. Conversation length, features, model and effort level all change how fast you use it up.

Usage-based Enterprise plans have no set limit. You are charged for what you consume, and you can track it under Settings > Usage.

Check usage in Claude Code

Run /usage inside a session. Per the command reference, /cost and /stats are aliases for it, and /stats opens on the Stats tab. It shows session cost, plan usage limits and activity stats. You can run it while Claude is responding.

The top of the screen is a Session block with token counts for the current session, like this example from the docs:

Total cost:            $0.55
Total duration (API):  6m 20s
Usage by model:
   claude-sonnet-4-6:  1.2k input, 5.3k output, 940.0k cache read, 50.0k cache write ($0.55)

Three things to know, all from Anthropic's cost documentation:

1. The dollar figure is computed on your machine from token counts at list price. It is an estimate. On Pro and Max, where usage is included in the subscription, the docs say the session cost figure is not relevant for billing.
2. The totals reset when /clear starts a new session. Before Claude Code v2.1.211 they kept accumulating across /clear.
3. On Pro, Max, Team and Enterprise, /usage also shows what counts against your plan limits: usage attributed to skills, subagents, plugins and individual MCP servers, plus recommendations. Press d or w to switch between the last 24 hours and the last 7 days. These figures are approximate and come from session history on that machine, so usage from other devices or claude.ai is not included.

If /usage says the usage endpoint is rate limited, it shows the last bars it loaded within the past 60 minutes. Press r to retry. A message like "rate limited until 3:14am" means Claude Code cannot fetch your limits until then, not that your work is blocked.

See what fills your context

/usage tells you what you have spent. /context tells you what is taking up room right now. It draws your current context usage as a colored grid and suggests fixes for context-heavy tools and memory bloat. Per the docs, running it is the way to see what MCP tool listings and other overhead cost you. This matters because Claude Code sends your whole conversation with each request, so a long session draws more from your limit than your last message suggests.

Keep it on screen with the status line

Run /statusline to configure a status bar that stays visible. The status line docs list the fields a script can read, including:

The rate_limits fields appear only for claude.ai Pro and Max subscribers (or behind a Claude apps gateway that sets a spend limit), and only after the first API response in a session.

Check usage credits

Run /usage-credits to manage usage credits, which let you keep working past your plan limit. On Pro and Max it opens Settings > Usage in your browser, where you can see your credit balance, this month's spend and your monthly spend limit. It is not available with API key authentication.

Check API usage in the Claude Console

If you use Claude through the API, Claude Code is billed per token and the authoritative numbers are in the Console. Anthropic's cost docs point to the Usage page. The Usage and Cost API docs also describe a Cost page at platform.claude.com/cost.

To pull the same data programmatically, the Usage and Cost Admin API has two endpoints: /v1/organizations/usage_report/messages for tokens and /v1/organizations/cost_report for USD. Per the docs, you need an Admin API key, and it is unavailable for individual accounts. Data typically appears within 5 minutes of a request finishing. Priority Tier costs are not in the cost endpoint, so use the usage endpoint for those.

On Amazon Bedrock, Google Cloud or Microsoft Foundry, spend shows up in your cloud billing console instead.

Why your usage may climb faster than expected

The cost docs list the usual causes: a long context re-sent with every request, cache misses after a break longer than the cache lifetime, scheduled tasks firing in an idle session, and subagents that each send their own requests. The documented habits are to run /clear between unrelated tasks, use /compact with instructions on what to keep, and match the model to the job. For a deeper look at why the dollar figure is only an estimate, see our piece on Claude Code usage cost.

Limits change. Check the pages linked above for the current numbers before you plan around any of them.