What are usage credits, and how do they let Claude Code continue after the limit?

Claude Code is Anthropic's coding tool, run in a terminal (the text window where you type commands). On a Pro or Max subscription its work is measured against a usage limit: the allowance of work the plan includes. The work is done in tokens, the pieces of text the model reads and writes, and Anthropic states the allowance as a multiple of another plan's allowance, with no token count attached. When that allowance runs out, Claude Code stops until the window resets. Usage credits are Anthropic's way to keep working before then. This page explains what they are, how to turn them on, what they cost, what changes in Claude Code once they are on, and what they do to the prompt cache. Every fact is Anthropic's own statement, read on 4 October 2026. The page is part of the guide to Claude Code usage limits.

What usage credits are

Anthropic's support article on usage credits, read on 4 October 2026, defines them: "Usage credits allow individuals subscribed to paid Claude plans (Pro, Max 5x, and Max 20x) to continue using Claude seamlessly after reaching their included usage limits. Instead of being blocked when you hit your session limits, you can switch to consumption-based pricing at standard API rates and continue your work without interruption" [1]. Consumption-based pricing means paying for each token, as an API customer does, instead of a monthly fee.

Two facts about the money. Usage credits "are billed at standard API rates", and the article points to Anthropic's pricing page for the rates, with the note that "price and plans are subject to change at Anthropic's discretion" [1]. And they "are charged separately from your paid Claude subscription and appear as additional charges on your bill" [1]. The page on a subscription or an API key for one developer sets out those per-token rates.

The article describes what happens at the limit once credits are on: "You'll see a notification that you've reached your session usage limit. If usage credits are enabled and you have funds available, you can choose to continue working. Your subsequent usage will be billed at standard API pricing rates. Your session limits reset every five hours as usual" [1]. Its question-and-answer section repeats the last point: the included limit "will reset every five hours once you reach it. Usage credits don't affect this reset timing" [1]. So credits add paid usage after the included allowance runs out, the allowance itself is unchanged, and when the window resets you are back inside the plan.

Anthropic says you get a warning first: "you'll see a clear notification when approaching and reaching your included usage limits, with a confirmation that you'll continue with usage credits" [1]. The windows themselves are the subject of the page on the five-hour window and the weekly window.

How to turn usage credits on

The usage credits article gives the steps, read on 4 October 2026 [1]:

1. Go to Settings > Usage in your Claude account on claude.ai.
2. Find the Usage credits section and click Enable.
3. Set up a payment method if you have not already.
4. Choose your spending preference. Click Adjust limit to set a monthly spend limit, or Set to unlimited for no limit.
5. Prepay for usage beyond the plan: click Add funds, enter the amount, and click Purchase.
6. Optionally turn on auto-reload, which buys more when your balance falls below a threshold you set.

Anthropic adds one cap on purchases: "There is a daily redemption limit of $2000" [1]. One restriction on where you can do this: "If you subscribed through one of the Claude mobile apps, you can only enable usage credits on the web version of Claude" [1].

From inside Claude Code, the /usage-credits command opens the same place. Anthropic's cost documentation says that for a Pro or Max subscriber it "opens Settings > Usage on claude.ai in the browser", where you "can turn usage credits on or off and check your credit balance, this month's spend, and your monthly spend limit" [2]. The command works after signing in with your claude.ai subscription through /login, and "isn't available with API key authentication" [2].

The spend controls Anthropic lists are a monthly spending cap, auto-reload settings, and usage alerts that notify you when you approach your spend limit [1]. Its own advice on the cap: "Start with conservative spending caps and adjust based on your needs" [1].

What Claude Code shows while credits are on

Once usage credits are on, /usage inside Claude Code gains a row. Anthropic's cost documentation, read on 4 October 2026, says what it shows on Pro and Max: "your spend for the current month, measured against your monthly spend limit when you have set one. When you haven't set a limit, the row shows Unlimited and no spend figure" [2]. With a spend limit set, "the row appears as soon as usage credits are on and shows 0% until you first spend usage credits" [2]. Before v2.1.236, /usage showed the row only on Pro and Max plans, and a row with a spend limit stayed hidden until you had spent something [2]. While usage credits are off, there is no row [2].

On claude.ai, the usage dashboard in Settings > Usage shows consumption as it happens, your month-to-date usage credit cost, and your usage history; Anthropic says the dashboard "clearly distinguishes between your included plan usage limits and any usage credit consumption" [1].

When the spend limit itself is reached, Claude Code shows a different message. Anthropic's error reference gives it as You've hit your monthly spend limit · raise it at claude.ai/settings/usage [3]. The condition, in Anthropic's words: "Your plan's included usage can't cover this request, and the usage credits that would otherwise pay for it have reached a spend limit" [3]. When one of the plan's windows is what ran out, the message also names the window's reset time, and access returns then without anyone raising the limit; before v2.1.239 it did not name that time [3]. On Pro and Max, when you reach the spend limit with credits still available, Claude Code prompts you to raise or remove the limit without leaving the terminal [2].

What usage credits do to the prompt cache

The prompt cache is a store, kept by Anthropic's service, of request text it has already processed, so the next request reuses it at a lower price. Cached text expires after a period with no requests, called the cache lifetime. Anthropic's prompt caching documentation, read on 4 October 2026, gives the lifetime by billing type. For the main conversation it is one hour on a Claude subscription within plan usage, and five minutes on usage credits, an API key or a cloud provider [4].

The reason is stated plainly: "Once you go over your plan's usage limit and Claude Code draws on usage credits, you are billed for that usage, so Claude Code drops the main conversation to the cheaper five-minute TTL" [4]. TTL is the time to live, another name for the lifetime. Anthropic says the one-hour lifetime bills cache writes at a higher rate [4], and on usage credits you are paying those rates.

The effect on you: after a pause of more than five minutes, the next request reprocesses the whole conversation instead of reading it from the cache. Anthropic lists this under why usage climbs in a long session: "your first message after a break longer than the cache lifetime misses the cache and reprocesses your full context. The lifetime is an hour on a subscription and drops to five minutes once you're drawing on usage credits" [2].

You can keep the hour. Set the promptCacheTtl setting, or the CLAUDE_CODE_PROMPT_CACHE_TTL environment variable, to 1h; each takes 5m or 1h and ignores other values, and both require Claude Code v2.1.242 or later [4]. The setting covers the main conversation only. Requests outside it, such as subagents, keep five minutes until you set subagentPromptCacheTtl as well, and Anthropic says it ignores a 1h in a subagent's own cacheTtl field while your subscription is using usage credits [4]. The page on the cache lifetime of five minutes or one hour covers when the longer lifetime is worth its higher write price.

Usage credits compared with waiting

	Usage credits off	Usage credits on
At the included limit	Claude Code blocks further requests until the reset time in the message [3]	A notification, then the choice to continue, billed at standard API rates [1]
What you pay	The subscription fee only [1]	The subscription fee, plus separate charges for usage past the limit, at API rates [1]
Session reset	Every five hours [1]	Every five hours, unchanged by credits [1]
Cache lifetime, main conversation	One hour within plan usage [4]	Five minutes, unless promptCacheTtl is 1h on v2.1.242 or later [4]
The /usage screen	No usage-credits row [2]	A row with this month's spend against your monthly limit, or Unlimited [2]
Where it is managed	Settings > Usage on claude.ai [1]	The same page, which /usage-credits opens from the terminal [2]
Upper limit on spend	The included allowance [3]	The monthly spend limit you set, or none if set to unlimited [1]; a $2000 daily redemption limit on purchases [1]
Can you turn it off	Off	Yes, at any time in Settings > Usage; you then have only the plan's included usage [1]

Two systems with similar names

Anthropic's support article on using Claude Code with a Pro or Max plan, read on 4 October 2026, describes a second path and says the two are "separate systems". In its words, Claude Code "presents options for continuing usage through API credits", while the Claude Console, Anthropic's website for pay-per-token accounts, "contains optional auto-reload settings for API credit management" [5]. If you choose API credits through Claude Code, "usage will be billed at standard API rates (distinct from Pro/Max Plan pricing)", and if auto-reload is on in your Console account, more credits are bought when the balance runs low [5].

That article's advice for staying inside the plan is to "decline the API credit option when presented" and to "allow your usage period to reset before continuing" [5]. To stop the API credit option appearing at all, it says to run claude logout, then claude login with only your Pro or Max credentials, and not to add Console credentials during login [5]. Its summary: "All transitions to API credit usage require explicit user consent" [5].

This page reports both articles as written. The usage credits article describes credits managed in Settings > Usage on claude.ai [1]; the Pro or Max article describes API credits tied to a Console account [5]. If you want paid usage to go through your claude.ai subscription's settings, the usage credits path is the one with the monthly spend limit and the /usage row described above.

Team and Enterprise plans

On a Team plan or a seat-based Enterprise plan, usage credits are turned on by an administrator for the organisation, and the /usage-credits command either opens the organisation's usage settings or sends a request to the administrators, depending on whether you have billing access [2]. Anthropic's release notes include the entry that introduced this, dated August 20, 2025: "A new setting allows Team and Enterprise plan users to purchase more usage to continue chatting with Claude or using Claude Code after hitting their usage limit" [6]. Spend limits at the organisation, group and individual level are covered on the page on spend limits for an organisation, a group and a person.

Where the credits can and cannot be used

Three further rules from Anthropic's documentation, read on 4 October 2026.

Usage credits apply across Claude and Claude Code: "Your combined usage across both interfaces counts toward your limits" [1]. They also apply to Research mode in the Claude app once you pass the included limit, and to all tokens processed, including documents in project files [1].

Some requests are paid only by usage credits, even with allowance left. Anthropic's error reference says the message Usage credits required for 1M context means the selected model uses the 1M-token extended context window "and your plan only includes it through usage credits"; after turning credits on, you restart Claude Code or start a new session, as the message says [3]. For the Fable models, Claude Code asks you to confirm before a Fable request bills usage credits, where your account requires that consent [3]. Anthropic's pricing page lists Fable on the Pro plan as available through usage credits, and on Max 5x and Max 20x as "50% of weekly limits", with an asterisk on the two Max cells [7].

Expiry: "In most cases, usage credits do not expire. However, in certain jurisdictions such as Japan, usage credits expire six months after purchase starting September 10, 2026", with an email seven days before expiry and the dates shown on the Usage page in Settings [1].

Our position

Turn usage credits on with a monthly spend limit before you need them, and leave the limit low. Anthropic's own advice is to start with conservative caps [1], and the /usage row then shows this month's spend against that cap from the moment credits are on [2]. Reveneau recommends also setting promptCacheTtl to 1h on Claude Code v2.1.242 or later, so a pause of more than five minutes during paid usage does not reprocess the whole conversation at API rates [4]. If you reach the included limit most days, the page on which plan includes what sets out the next tier, which Anthropic's Pro or Max article names as the answer to consistent limits [5].

Reveneau is an AI software development consultancy, and all of its code is written by AI, so usage past a plan limit is a running cost of every Reveneau build. Reveneau is independent of Anthropic. Every figure on this page is Anthropic's own statement about its own product, read on 4 October 2026.