Skip to content
Menu

Guides

How to find out who on your team went past their plan, and what it cost

Checked on 3 November 2026 against the sources below.

On a Claude Team or Enterprise plan, a member can only spend past their seat if usage credits are turned on. With credits off, a member who reaches their 5-hour or weekly limit stops until it resets and costs nothing extra. With credits on, an Owner or Primary Owner sees each member's month-to-date usage-credit spend in Organization settings > Usage, and a spend report lists it by user and by model, a day behind, billed at API rates; the Owner can set a monthly spend limit for the whole organisation and for each member. On Max plans there is no team view: each person's own Settings > Usage shows their usage-credit spend and limit, and the company sees the card statement. None of Anthropic's screens show what happened around the limit, which session or repo it was, what the person was doing, or whether anything shipped. OpenReason Intel, a dashboard for Claude Code and Codex usage, shows that: each over-limit period with the person, session, repository and model around it, the turns underneath, the cost with its qualifier, and the commits or pull requests recorded in it.

The two meanings of "past the plan" come first, then the controls Anthropic gives you, then what OpenReason adds and what it does not do.

Two things "past their plan" can mean

The first is hitting the window. Every Team and Enterprise seat has a 5-hour and a weekly allowance, and the Claude Code docs say a member's usage "draws from a per-seat allowance that resets on a rolling five-hour window and a weekly window". When it is used up, the member sees "You've hit your session limit" or "You've hit your weekly limit" and waits for the reset. This costs the company nothing beyond the seat. It costs the member their afternoon.

The second is spending past it. Anthropic's usage credits (called extra usage until recently) let a member keep going after the window fills, "billed at standard API rates". This is the overage, and it is the only part an admin sees in dollars, because "usage inside the seat allowance isn't metered in dollars".

So "who went past their plan" has a short answer on Anthropic's side: nobody, unless you turned usage credits on. The longer answer, who keeps hitting the window and what that is costing in stalled work, is not on any vendor screen.

On a Team or Enterprise plan: what the Owner sees and sets

From the Team usage-credits article, dated 4 September 2026:

  • Turning it on. An Owner or Primary Owner enables usage credits under Organization settings > Usage. Both Standard and Premium seats can use them once enabled.
  • The organisation cap. The Owner can "either input an amount and click 'Set spend limit,' or click 'Set to unlimited'" for the month.
  • Per-member caps. "Owners and Primary Owners can also set individual monthly spend limits for each member", in the Spend limits by user section.
  • Who spent what. The members list has "a MTD Spend column, so you can track members' usage patterns and optimize seat assignments".

The spend report in usage analytics adds the detail: a CSV with each user's email, the model, request count, prompt and completion tokens, and net spend. "Spend data refreshes daily and has a one-day delay."

A member without billing access cannot turn credits on for themselves. The Claude Code docs' table for /usage-credits says that for a "Team or Enterprise member without billing access" the command "asks you to confirm, then sends a request to your organization's admins". When a member then hits a cap you set, they see one of three messages, "You've hit your individual spend limit", "org's monthly spend limit" or "team's shared budget", and the docs tell the admin to raise the limit the message names under Organization settings > Usage, or let the member wait for the plan reset.

On Max plans the company pays for

If your engineers each have a Max plan on a company card, there is no Owner and no report. The Claude Code docs: "On an individual Pro, Max, or Ultra plan you have no organization to manage, so track your own usage-credit spend". Each person turns usage credits on or off under their own Settings > Usage, sets their own monthly spend limit with "Adjust limit", and sees this month's spend in /usage. The company's view is the card statement, one line per person per month, with the usage-credit charges on it.

The practical arrangement most teams land on is to ask each person to set a monthly spend limit and to leave credits off unless they need them. It is a request, not a control: the person can change it, and you will not know until the statement arrives.

The controls you actually have

Put together, Anthropic gives a team these levers and no others:

  • Leave usage credits off. The window is then a hard stop and the overage is zero. The cost is the time people spend waiting for a reset.
  • Turn them on with caps. An organisation-wide monthly limit, and a limit per member for anyone who should have a different one.
  • Choose seats. A Premium seat has "5x more usage than standard seats" on claude.com/pricing, at $125 a month against $25; for someone who hits the Standard window every day it can be cheaper than their credits.
  • Let Claude Code wait. From v2.1.234 it can pause at the limit and continue the task after the reset, and the managed setting autoContinueAtUsageLimit sets that for a whole fleet, per the costs docs.

What you cannot get from these is the why. The spend report says a person spent $40 on usage credits last Tuesday; it does not say that a session had been open since Monday and was sending a day of context with every request, or that three agents ran at once, or that nothing was merged at the end of it. That is the gap the next section is about.

See who went past their plan, and what was happening, in OpenReason Intel

OpenReason Intel records each person's 5-hour and weekly windows from their own machine, Claude Code's from its status line and Codex's from its session file, and keeps the history. Its Incidents view lists each over-limit period with the person, the session, the repository and the model, and under it the quota and overage charts with the turns that made them. From there you can inspect the work, the requests, the context size and the tools around the incident, which is the question "what used it all" needs answered.

  • Cost. An incident carries a cost figure with its qualifier: vendor-reported when the team has turned on the option that lets or-otel fetch each person's usage-credit spend from Anthropic with their own Claude Code sign-in, otherwise allocated from the vendor's amount or estimated at list price. It is not an invoice.
  • Diagnostic tags. An incident is tagged with the patterns found around it, such as long context, a cold restart or a parallel burst. They are patterns to investigate, not proof of the cause or of waste.
  • Recorded output. The commits, pull requests or merges recorded in the period sit beside it. A period with none does not mean no useful work was done.
  • Per person. Each person's page shows their 5-hour and weekly history, active days, incidents, machines and versions, and recorded work, so one bad week and a habit look different.

What it does not do. It does not cap anyone, stop a request or send an alert; the caps stay in Anthropic's settings where they are, and OpenReason shows you what happened inside them. Because the data is on each machine and in the dashboard, a team with a coding agent can build its own warning on top. A figure beside a name is evidence for a conversation, not a verdict on a person; the turns under the chart are there so the conversation starts from what the work was. Each engineer sees the same about themselves and nothing about a teammate.

It reports Claude Code and Codex, from Mac, Windows and Linux machines, with one command each from the or-otel quickstart. It is free under ten engineers and $20 an engineer a month after that, on the pricing page; sign-up takes an email address or a Google account, no card.

What to do next

  1. Read next

  2. See what used up your limit

    Session Forensics is a free page that reads the Claude Code and Codex session files already on your computer, in your browser. It lists the limits you hit and the turns that used each one up. There is nothing to install and no sign-up, and your files are never uploaded. For Claude Code its 5-hour figure is an estimate, because Claude Code's files do not record the meter.

    Would you rather have your own agent do it? Paste this prompt into Claude Code or Codex.

    Open Session Forensics (opens in a new tab)

  3. Get an email when the limits change

    Get an email when Claude Code or Codex limits change. One email per change, at most one a week, from OpenReason. Unsubscribe from any email in one click.

  4. Track it over time

    OpenReason Intel keeps a record of your Claude Code and Codex 5-hour and weekly windows from every machine you use, with each turn's estimated share of the window and what it cost once the window was full. It is free for small teams: sign up with Google, no card, then run one command on Mac, Windows or Linux.

    It shows you what happened; it does not stop a session as you work, and it cannot change your vendor's limits.

    Sign up free with Google

Questions people ask

How do I find out who on my team went past their plan and what it cost?
On a Team or Enterprise plan, if usage credits are on, the Owner sees each member's month-to-date usage-credit spend under Organization settings > Usage and in the spend report, by user and model, a day behind. With credits off, nobody can spend past their seat. OpenReason lists each over-limit period with the person, session, repo and model, the turns under it, the cost with its qualifier, and what was recorded as shipped.
How do I manage overages on Claude Team or Max plans for a team?
On a Team plan, an Owner turns usage credits on and sets a monthly spend limit for the organisation and for each member, or leaves them off so the window is a hard stop. On Max plans each person manages their own credits and limit and the company sees only the card statement. OpenReason shows the history and the turns around each over-limit period so you can see what the overage bought; it does not cap or alert.
How do I see what my team spends on Claude Code?
Seat price times seats, plus the usage-credit spend in the spend report, which appears once credits are on. On the API, the Console dashboard shows estimated spend per member. OpenReason shows seat commitments and variable spending together per person; the guide on what your team spends on Claude Code covers each case.
How do Claude Code's 5-hour and weekly limits work?
The 5-hour window starts on your first message and resets in full five hours later; the weekly limit resets at a fixed time each week assigned to the account. Each Team seat has its own pair. The guide on how Claude Code's limits work has the detail and the plan comparison.

Sources

Every figure on this page was read from these pages on 3 November 2026. Limits and prices change; if a vendor page now says something different, the vendor page is right and this one is out of date.