Usage credits let a paid Claude plan keep working after the included allowance runs out, billed separately and capped by spend limits you control. Anthropic's documentation says Pro and Max users can set a monthly limit, Team and Enterprise admins can set limits by organization, group or member, and the prompt cache lifetime shortens while credits are in use.
What they are
Anthropic's Pro plan help article says Pro subscribers can enable usage credits to keep working beyond their included limits, and that limit resets can also restore session or weekly limits to full. The Claude Code cost documentation, read on 19 October 2026, describes a command that opens the right settings page depending on your role:
- Pro or Max: opens your usage settings, where you can turn credits on or off and see your balance, this month's spend and your monthly spend limit.
- Team or Enterprise with billing access: opens the organization's usage settings.
- Team or Enterprise without billing access: asks you to confirm, then sends a request to your organization's admins.
The help center article on extra usage has the rates and rules. This post did not read it, so check that page for how credits are priced before you turn them on.
Where the spend shows up
The same documentation says the usage screen shows a credits row while they are on. On Pro and Max it shows spend for the current month against your limit, or "Unlimited" if you set none. On Team and Enterprise it shows your own spend against any limit that applies to you. The spend report in organization analytics, updated daily, covers usage credit spend and appears once credits are turned on.
Set a limit before you need one
An unlimited credit setting on a personal card is an open tap. Set a monthly limit that you would accept as a bad month. For organizations, the page describes limits at three levels: the whole organization, a group, or an individual member. Messages mention the name of the limit hit, such as an individual limit, an organization monthly limit or a team shared budget, and tell admins which setting to raise.
A side effect worth knowing
The Claude Code documentation says the prompt cache lifetime is an hour on a subscription but drops to five minutes once you are drawing on usage credits, and that you can choose the time to live yourself to keep the one hour behavior. A shorter cache means more cache misses after a pause, and each miss reprocesses the full context. See why agent bills balloon.
Credits, a higher tier or the API
- Occasional overflow: credits with a low cap.
- Regular overflow every week: compare a higher tier, as in Claude plans compared.
- Automation: use the API with a workspace limit. See API spend caps.
Quick check
Is waiting for a reset free?
Yes, waiting costs nothing. Anthropic notes that newer Claude Code versions can wait and continue an interrupted task automatically after the reset.
Who can turn credits on for a team?
Per the documentation, members with billing access manage them in organization settings, and other members send a request to admins.
Sources
Last reviewed: 19 October 2026.