Key takeaways
- As of November 2026, Pro has a five-hour and a weekly limit, Max gives 5x Pro on both, and verified Ultra accounts get 10x Pro weekly with no five-hour limit.
- Each person's allowance covers chat, Cowork and Claude Code together, and usage counts against the five-hour and weekly windows at the same time.
- UsageAtlas, a free desktop tracker, lets each team member log a normal week of Claude usage before you choose seat types.
Claude usage limits come in two layers on most subscriptions: a session window that refills five hours after it opens and a weekly allowance that resets at a fixed time for each account. Anthropic sizes those layers as multiples of the Pro plan rather than as message counts, so the price list alone will not tell a founder how much work a seat can carry.
That makes buying Claude for a small team a measurement problem. This is what Anthropic's pricing page and help center say as of November 3, 2026, and how to turn it into a budget.
How each plan limits usage
| Plan | Price | How usage is capped |
|---|---|---|
| Pro | $20 a month, or $17 a month billed yearly | Five-hour session limit and a weekly limit |
| Max | $100 a month | 5x Pro every five hours and every week, plus a separate weekly Fable limit |
| Ultra | $200 a month | 10x Pro every week, no five-hour limit once the account is verified, separate Fable limit |
| Team, standard seat | $25 a month, or $20 billed yearly | Per-seat five-hour and weekly windows, more than Pro |
| Team, premium seat | $125 a month, or $100 billed yearly | 5x a standard seat |
| Enterprise | $20 a seat billed yearly, 20 seats minimum, plus usage at API rates | Consumption billing with spend limits set by admins |
| API through the Console | Per token | Per-minute rate limits and a monthly spend cap by usage tier |
Anthropic adds that it may cap usage in other ways, such as monthly caps, and does not promise uninterrupted service. Plan for headroom.
One allowance per person, many places to spend it
Each person's allowance covers chat, Cowork and Claude Code together. Anthropic lists the drivers: conversation length, attachments, the model, the effort level, tools such as web search, and multi-step tasks that run code or create files.
The two windows also stack. Work counts against the five-hour window and the weekly one at the same moment, so a developer who pushes hard through Monday can be out of weekly allowance by Wednesday even though each session felt fine.
Overflow without surprise bills
When a window runs out, work stops until the reset unless usage credits are on, which bill the extra at standard API rates.
- On Pro, Max and Ultra, the account owner turns them on in Settings > Usage, sets a monthly spend limit and can add auto-reload.
- On Team and seat-based Enterprise, admins turn credits on and can set spend limits for the organization, a group or a single member. A spend report with CSV export shows estimated spend per user and per model once credits are on.
- The monthly API credits bundled with Max and Ultra cannot be spent in Claude Code or the Claude apps.
The API is a separate world. There is no five-hour or weekly window; instead each organization gets requests-per-minute and tokens-per-minute limits that refill continuously, plus a monthly spend cap. One trap to brief the team on: if ANTHROPIC_API_KEY is set on a laptop, Claude Code bills that key instead of the person's subscription.
Measure a normal week before you buy seats
Usage inside a seat allowance is not metered in dollars, so the admin console will not tell you whether a standard seat is enough. The cheapest way to find out is to have each person track one ordinary week.
UsageAtlas is the tool we would hand them. It is a free, open-source desktop app for Windows, macOS and Linux that shows each person's Claude five-hour and weekly windows, reset times and Claude Code history, with Codex and Cursor beside it. No account is needed for local tracking, and it does not read prompts or responses. After a week, anyone who hit the weekly limit more than once is a candidate for a premium seat or Max; everyone else can stay on the base tier. Get UsageAtlas here.
UsageAtlas runs on each person's own computer, though. If you need one org-wide report, Anthropic's spend report on Team or the Enterprise Analytics API is the better fit.
Frequently asked questions
Do Team seats have Claude usage limits?
Yes. Each Team member draws from a per-seat allowance with a rolling five-hour window and a weekly window. A premium seat has five times the usage of a standard seat.
Can a team keep working after hitting a limit?
Yes, if an admin turns on usage credits. Extra work is billed at standard API rates, and admins can set spend limits for the organization, a group or one member.
Does the Claude API have weekly limits?
No. The API uses per-minute request and token limits that refill continuously, plus a monthly spend cap set by your usage tier.
Sources
- Plans and pricingAnthropic
- How do usage and length limits work?Anthropic
- What is the Ultra plan?Anthropic
- Manage usage credits for paid Claude plansAnthropic
- Manage costs effectivelyAnthropic
- Rate limitsAnthropic
- UsageAtlasUsageAtlas