Claude Code usage limits and usage credits on Team and Enterprise plans, explained

How the per-seat session and weekly limits work, what usage credits are, what changed on September 14, 2026 and what changes at renewal, what drains a seat's weekly allowance, and what an engineering organization should check before the next reset.

Updated September 13, 2026

On September 14, 2026 the weekly limit on every Claude Team and seat-based Enterprise seat changed. From May 13 through September 13 it was 50% higher than the published standard² . From September 14 the standard is 25% higher than it was before that promotion³ , which the vendor describes as a 17% reduction compared to the day before⁴ ⁵ . A Standard seat is $25 a month, or $20 billed annually, and a Premium seat $125, or $100 billed annually¹ ¹¹ . The next weekly reset is the first one on the new standard.

fig. 3b — weekly limit per seat, indexed: pre-promotion standard = 100 ■ standard from September 14, 2026

Period	Weekly limit per seat, indexed	What the vendor says
Before May 13, 2026	100	The published standard
May 13 to September 13, 2026	150	“your weekly usage limit in Claude Code is 50% higher"²
From September 14, 2026	125	“permanently raising standard weekly limits in Claude Code by 25%“³ , which “works out to a 17% reduction” compared to the promotional level⁴

The index is ours. The vendor’s statements give the changes as percentages, and this page states them the same way, so no absolute figure for a weekly limit appears here. The arithmetic from the vendor’s two percentages is 100 × 1.5 = 150 during the promotion and 100 × 1.25 = 125 from September 14, and 125 is 16.7% below 150, the figure the vendor states as 17%⁴ .

Usage limits and usage credits

Two units, and only one of them is money:

A usage limit is the allowance included in a seat: a rolling five-hour window and a weekly window⁷ , sized by seat tier. A Standard seat carries 1.25 times the Pro plan’s usage per session and a Premium seat 6.25 times¹ . Usage of every Claude surface counts against the same limit⁶ , it is metered in tokens¹³ , and what it costs depends on “the length and complexity of your conversations, the features you use, which Claude model you’re chatting with, and the effort level you’ve selected"⁶ . Weekly limits reset at a fixed time each week that is assigned to the account, and a member receives the full weekly allowance each cycle¹ . The allowance is the member’s own: usage limits on Team plans are per member, not applied to the team as a whole¹ .
A usage credit, formerly called extra usage, is what a member on a Standard or Premium seat spends to continue working with Claude, Cowork and Claude Code after reaching the included limit⁸ . Usage credits are off until an Owner or Primary Owner turns them on, they bill at standard API rates, and they run against monthly spend limits set for the organization, per seat tier and per member⁸ . Team organizations prepay them; seat-based Enterprise organizations are billed at the end of each month⁸ . A member who exceeds their spend limit cannot use Claude, Cowork or Claude Code again until the next billing period⁸ .

Usage inside the allowance is not metered in dollars⁷ . The dollars start at the wall.

Which plan you are on

Plan	What a seat includes	What governs consumption past it
Team, Standard or Premium seat	The per-seat session and weekly windows, sized by seat tier and belonging to the member¹	Usage credits, off until an owner enables them, at standard API rates against monthly spend limits⁸
Seat-based Enterprise, legacy Standard or Premium seat	The same per-seat windows⁷	Usage credits, billed at the end of each month⁸ . The plan moves to the usage-based seat at the next contract renewal¹⁰
Usage-based Enterprise, the single seat	Access only, with “no per-seat usage limits and no included token allowance"⁹	Every token at standard API rates, under organization and individual spend limits⁹ ¹⁰

Individual Pro and Max plans are not covered on this page.

How the seat model is changing

fig. 3a — one meter per seat, nothing pooled ■ at the weekly wall

Under the per-seat model, one member’s heavy week is one member’s problem, and one member’s quiet week is nobody’s gain. Four changes in 2026 moved the walls without changing the shape.

May 13, 2026. The weekly limit in Claude Code became 50% higher on Team seats and on legacy seat-based Enterprise seats, with no action required. Free plans and consumption-based Enterprise seats were not included² . The promotion was extended and ran through September 13, 2026² .
July 20, 2026. The promotion under which Fable 5 was included in a Standard seat’s limits ended on July 19, 2026 at 11:59:59 PM PT. On Team Standard seats, Fable 5 and Fable 5.1 are not included in the plan’s usage limits and run on usage credits, with a one-time credit issued when Fable 5 moved in July 2026¹² . On Premium seats, up to 50% of the weekly limit can be used on Fable models at no extra cost, and they use it faster than other models¹² .
At your next contract renewal, on Enterprise. Legacy Chat, Chat plus Claude Code, Standard and Premium seats transition to the single usage-based Enterprise seat¹⁰ : a seat price plus usage at API rates, at $20 per seat¹¹ , with no per-seat usage limits and no included token allowance⁹ . The Register reported the shift on April 16, 2026¹⁵ .
September 14, 2026. Weekly limits in Claude Code returned to their standard levels² , with the standard raised 25% above where it stood before May 13³ , which is 17% below the promotional level⁴ ⁵ .

Three details that matter more than the dates:

Usage credits are off by default⁸ . Doing nothing is a decision that every seat stops at its wall.
Switching models does not restore access. The weekly limit is shared across all models, so a member who has hit it cannot get back to work by choosing a cheaper one⁷ .
The reset does not wait for anyone. The reset day and time stay the same regardless of when a member starts using Claude, and each cycle begins with the full allowance¹ . Allowance a member did not use by the reset is not theirs the following week, and it was never anyone else’s.

Why this month will look different

A team whose working pattern fit inside the promotional weekly limit has 17% less room from September 14⁴ . On a seat-based plan the dollars appear only past the wall, as usage credits, so the invoice moves with the number of members who cross it and how far.

17%

less weekly allowance per seat from September 14 than during the promotion, per the vendor

$13/day

average cost per developer per active day across enterprise deployments, per the vendor

$150–250/mo

average cost per developer per month across enterprise deployments, per the vendor

Quoted figures, not modeled ones.⁴ ⁷ The averages are stated across enterprise deployments without a plan breakdown, and the vendor adds that costs stay below $30 per active day for 90% of users⁷ . The other 10% are the tail, and on a per-seat plan the tail is exactly the set of members who reach the wall.

What actually drains a weekly allowance

fig. 3c — one week, one seat each ■ blocked ■ resets unused

Every turn sends the conversation so far, the project context and the new prompt, and the conversation is the part that grows fastest¹³ . Opus costs several times more per turn than Sonnet¹³ . The vendor’s consumption guide ranks Claude Code and Cowork above chat in token intensity, because each coding session carries system prompts, file context, tool calls and multi-turn reasoning, and agentic workflows generate significant intermediate token usage¹⁶ . So the seats that hit the wall are the seats running agents, and the analytics dashboard has a panel for exactly that: spend concentration¹⁴ . The consumption guide tells administrators to review individual usage reports monthly for outliers of both kinds, members consistently hitting their cap and members consuming very little¹⁶ .

“Weekly limit (all models): 100% used”, three days after the seat was provisioned. One reported case, a Team Premium seat, March 2026¹⁷

Six things to check before the next reset

Decide whether usage credits are on. They are off by default⁸ . Off means every seat stops at its wall. On means the organization limit and the per-member limits need setting before the first member crosses, because a member past a limit is blocked until the next billing period⁸ .
Re-forecast against the new standard, not the promotion. A pattern that fit in August has 17% less room now⁴ . Members who reached the wall on the promotional limit will reach it earlier on the standard one.
Check the default model and effort level. Opus costs several times more per turn than Sonnet¹³ , effort level is one of the four things the vendor says usage depends on⁶ , and on Standard seats Fable models run on usage credits rather than inside the limit¹² . A cheaper default is one of the few levers that works without changing anyone’s behavior.
Look at the distribution, not the average. Cost analytics in Settings shows spend concentration and spend by model¹⁴ , and the spend report export lists every member’s spend per model over up to 90 days¹⁴ . If you would rather work from that export, the Claude spend report we publish reads it in your browser and draws the distribution, including who is at or over a per-member limit you type in. Nothing is uploaded, and it needs no token or account. On Team and seat-based Enterprise plans the export covers usage-credit spend only, and it appears once usage credits are turned on⁷ .
Know your renewal date if you are on a legacy Enterprise contract. The seat-based plan transitions to the usage-based single seat at the next renewal¹⁰ , and from then on there is no included allowance to forecast against, only spend limits and the invoice⁹ .
Decide what happens when someone hits the wall on Thursday. The reset is at a fixed time¹ . A member can wait for it, or request usage credits and wait for the daily email that puts the request in front of an administrator⁸ . Either you have a process, or your engineers have a workaround.

The part nobody has solved

Every native control, the usage credits switch, the organization, seat-tier, group and per-member spend limits, the request queue, is a ceiling. Ceilings stop spending, and none of them moves unused allowance to the member who ran out. Both results are true in the same organization, in the same week: allowance resetting unused on one seat, a blocked engineer on the next. We wrote up what each control does and where it stops for anyone choosing a limit this month.

fig. 2e — the missing route (cf. fig. 1, front page)

We are building Spillway Ops for that missing route: allowance that would otherwise reset unused flows to the members who ran out, so spend falls without anyone losing a tool. It reads usage metadata and sets per-member limits through vendor admin APIs with no repository access, no prompt contents, no code. The Claude Code connection is in development. If the first week on the new standard is about to be a conversation at your company, we would genuinely like to hear how it goes.

Drop your email for early access notice and product updates in between.

Sources

Claude Help Center What is the Team plan?
Claude Help Center Claude Code May–August 2026 weekly limits promotion
Claude Developers on X Starting September 14, we’re permanently raising standard weekly limits in Claude Code by 25%
Claude Developers on X Compared to today, this works out to a 17% reduction in weekly limits on Claude Code
BleepingComputer Anthropic is cutting Claude Code’s current weekly limits by 17%
Claude Help Center How do usage and length limits work?
Claude Code Docs Manage costs effectively
Claude Help Center Manage usage credits for Team and seat-based Enterprise plans
Claude Help Center What is the Enterprise plan?
Claude Help Center How am I billed for my Enterprise plan?
claude.com Pricing
Claude Help Center Claude Fable models on your plan
Claude Help Center Models, usage, and limits in Claude Code
Claude Help Center View usage analytics for Team and Enterprise plans
The Register Anthropic ejects bundled tokens from enterprise seat deal
Claude Help Center Claude Enterprise consumption guide
GitHub, anthropics/claude-code Team Premium seat weekly limit stuck at 100% after only 102M tokens (issue 41001)

Last updated September 13, 2026.