Claude cloud session credits: how they're used
A cloud session runs Claude Code in the cloud, on Anthropic's machines by default. See which balance it uses, what the credits are, and when they run out.
Why a cloud session spends a credit when you already pay for a plan
A cloud session credit is a free bonus balance that a cloud session draws down before it touches the usage your plan already includes. Anthropic granted it to eligible individual Pro and Max subscribers as a promotion, so it is not money you paid in. A cloud session is a Claude Code session that runs in the cloud instead of on your laptop: by default on machines Anthropic manages, or on an organization's self-hosted environment when the organization routes sessions there. It still belongs to your subscription, and Anthropic's documentation is explicit that there is no separate charge for the virtual machine (VM) the session runs on. What the credit changes is the order in which the meter is read: while a credit balance is attached to your account, cloud sessions spend that balance first, and only after it is empty do they start counting against the same pool your terminal sessions and your chats already share. The promotion's credit expired on 4 November 2026, so cloud sessions now count against that shared pool from the start.
So the answer to "why is this charging a credit when I already pay for Pro, Max, or Ultra" is that the credit is an extra pool sitting in front of your plan, not a replacement for it. Nothing about your subscription changed. A second meter appeared ahead of it.
Is this session running on your machine or in the cloud?
Everything else depends on this distinction, and the interface does not always make it loud. A session runs in the cloud when you start it from claude.ai/code in a browser, from the Code tab in the Claude mobile app, from the Desktop app with Cloud selected instead of Local, from the terminal with claude --cloud, or from a scheduled routine. A session runs on your own machine when you start it in your terminal, in your IDE, or in the Desktop app with Local selected.
One case looks like an exception and is not. Remote Control lets you watch and steer a session from your phone or browser while that session executes on your own machine. It is a local session, so it never touches a cloud credit, because the work is running on hardware you already own.
Cloud sessions are available on Pro, Max, Ultra, and Team plans, and to Enterprise users with premium seats or with Chat + Claude Code seats. If custom roles govern your access, you also need a role with the cloud sessions permission to start sessions on Anthropic's infrastructure.
Where does a cloud session's usage come from?
From your account's normal pool, once any credit in front of it is gone. The cloud sessions documentation states it plainly: "cloud sessions share rate limits with all other Claude and Claude Code usage within your account. Running multiple tasks in parallel consumes more rate limits proportionately." The same page adds that "there's no separate charge for the cloud VM."
Read the second sentence twice, because that is the one that surprises people. Each claude --cloud command creates its own session with its own context window. Start four tasks at once and four sessions send their own requests at the same time, so the draw on your account is roughly four times what one session would use. The VM is free. The tokens are not.
The exceptions are features that bill to usage credits instead of your plan, covered in the next section.
The word "credit" means two different things
The interface uses one word for two separate mechanisms, and that overloading is most of the confusion behind this question.
Usage credits are the permanent mechanism. They are a prepaid balance you fund yourself in Settings > Usage, available on Pro, Max, and Ultra. For most work they do nothing until you reach one of your plan's limits: Anthropic describes them as kicking in "instead of being blocked when you reach one of your plan's limits". Some model and feature choices are the exception and draw on them from the start, as the next paragraph explains. After that point, if credits are on and you have funds, your work continues and usage is "billed at standard API pricing rates". API means application programming interface, and standard API rates are charged per token, so this is metered spending rather than an allowance. You can set a monthly spend cap, and /usage shows the month's spend against it. The cap is not the only stop: Anthropic says usage continues "until you reach your monthly spend limit, or your balance runs out with auto-reload off", so a balance you funded once can run dry before the cap is reached. Usage credits cover both Claude conversations and Claude Code. For the whole picture of that balance, including the second unrelated sense the word carries, read what usage credits are and where the word gets overloaded.
Some features spend usage credits even while plan usage remains. On Pro, Max, Ultra, Team, and Enterprise, fast mode is "available via usage credits only and not included in the subscription rate limits", and the fast mode docs add that it "draws directly from usage credits, even if you have remaining usage on your plan". Until credits are on, /fast refuses with "Fast mode requires usage credits". Fast mode works in cloud sessions too. Ultrareview, which runs as a cloud session, gives Pro, Max, and Ultra three free runs per account, then bills each review to usage credits. Model choice matters as well. On Pro, Anthropic's Fable article says Fable models normally run on pay-as-you-go usage credits from the start, because they are not part of the plan's usage limits. A Fable 5.2 launch offer changes that for one week: from November 9 through November 16, 2026 (11:59 pm PT), Pro can "use up to 50% of your weekly usage limits on Fable models at no extra cost", and the offer covers "every Fable model, including Fable 5.2, Fable 5.1, and Fable 5." Anthropic adds: "After you use that 50%, or once the week ends, Fable usage is billed to usage credits if you have them turned on." On Max and Ultra, Fable can use up to half of your weekly limits at no extra cost, then continues on usage credits. That is their regular plan allowance. The launch offer's included week applies only to Pro and standard Team seats. Eligible Pro and standard Team seats also get one Fable reset, which expires November 16, 2026 at 11:59 pm PT. Max, Ultra, premium Team seats, and premium seats on legacy seat-based Enterprise plans get no included launch week; eligible accounts receive one to three one-time Fable resets, depending on subscription length, which expire March 8, 2027 at 11:59 pm PT. These are the cloud work that costs money on top of your plan.
A cloud session credit is a promotional bonus, not a prepaid balance. Anthropic's help center describes it as a bonus credit of $100 on Pro and $250 on Max, and says "Cloud sessions are included in your Pro or Max plan, and this credit adds free usage on top of your regular plan limits." It was granted rather than purchased, and it covered only cloud sessions you started yourself: Projects and Routines were excluded even though routines run as cloud sessions, and so were chat, Cowork, and Remote Control sessions. It was applied automatically when a cloud session started. While it lasted, cloud session usage was paid from it and did not count toward your plan's limits. One wrinkle on Pro: the terms said "Your bonus credit is used first, but Fable will draw on your usage credits immediately after the bonus credit is spent." Selecting Fable on Pro required usage credits to be turned on with a balance, even while the bonus credit remained.
The monthly API credits now included with Max ($100 a month) and Ultra ($200 a month) are separate from both balances above. They go to a linked Claude Console organization and cannot pay for interactive Claude Code or usage credits in Claude, Claude Code, or Cowork.
Here the honest answer is that the two are documented in different places, so check your own account rather than trust any summary, including this one. As of 10 November 2026, the cloud sessions page at code.claude.com covers rate limit sharing and says nothing about a promotional credit, while the usage credits help article covers the prepaid balance and says nothing about cloud sessions. The promotion has its own help center article. It was open to individual Pro and Max subscribers whose subscription was active when it began on 23 September 2026; Team, Enterprise, free trial, and past-due accounts were not eligible, and neither were sign-ins through Amazon Bedrock, Google Vertex AI, or an API key. The claim deadline was 7 October 2026, and the credit expired on 4 November 2026 at 11:59 PM PT. Until then, the remaining balance showed in the usage menu.
What happens when the pool is empty?
Three different things, depending on which pool emptied.
When a promotional cloud session credit ran out, nothing stopped. Cloud sessions went back to counting against your normal plan limits, except Fable on Pro, which went straight to usage credits because Fable normally sits outside the Pro plan's limits. For one week that rule has an exception: from November 9 through November 16, 2026, the Fable 5.2 launch offer includes every Fable model (Fable 5.2, Fable 5.1, and Fable 5) in Pro's weekly limits, up to 50% of them, so Pro Fable usage draws on that included allowance first and moves to usage credits only after the 50% is used or the week ends. An earlier promotion that counted Fable 5 toward Pro's weekly limits ended July 19, 2026, and never applied to Fable 5.1 or 5.2. This was the case worth watching, because the credit was the reason cloud work felt free, and the switch back was silent.
When your plan's included limit is reached and usage credits are on with funds available, work continues and is billed per token at standard API rates until you reach your monthly spend limit, or until your balance runs out if auto-reload is off. Auto-reload, which buys more credit when the balance falls below a threshold you set, closes the second gap. At the spend limit Claude Code reports a message naming the limit it hit, such as "You've hit your monthly spend limit", and on Pro, Max, and Ultra you can raise the cap in Settings > Usage, which /usage-credits opens in your browser.
When your plan's limit is reached and usage credits are off or unfunded, new work stops. In eligible terminal, IDE, and Desktop Code sessions running Claude Code 2.1.277 or later, a small, capped wrap-up allowance is rolling out that lets an already running response reach a stopping point after a five-hour limit is reached; it still counts toward the weekly limit. Verified Ultra accounts have no five-hour limit; accounts not yet verified have a temporary limit. You see "You've hit your session limit" or "You've hit your weekly limit", along with the time that window resets. Those limits cover every model, so /model will not rescue you. A model-specific message behaves differently: after "You've hit your Opus limit", switching to another family with /model sonnet does keep you working. When that limit is the weekly Opus allowance, what it means when Opus priority access is used up covers when Opus comes back and whether credits or a Max upgrade restore it sooner. The options at the moment a limit lands and how the reset windows are counted are both easier to read before you need them than during.
Nothing on this path quietly downgrades your model or trims your context to keep going. The meter either has something to draw on or it stops. If the message that stops you mentions Claude's safeguards instead of a limit, credits and reset times are not what's in play, and what a safeguards message means for the session and your account covers the next step Anthropic names.
Where do you see what a cloud session consumed?
Two places, and one of them misleads you if you read it wrong.
Settings > Usage on claude.ai is the authoritative view. It holds your plan usage, whether usage credits are on, your balance, this month's spend, and your monthly spend limit. From a terminal, /usage-credits opens that same page in your browser.
/usage inside a session shows your plan usage bars, and on Pro, Max, Ultra, Team, and Enterprise it adds a breakdown of where recent usage went: skills, subagents, plugins, and individual MCP servers, each as a percentage of the total. It also flags behaviors that account for a large share of recent usage, such as long context or cache misses. Press d or w to switch between the last day and the last week.
Now the trap. Those day and week figures are computed from session history on the machine you typed the command on. The documentation says so directly: "usage from other devices or claude.ai is not included." So running /usage in your terminal after an afternoon of heavy cloud work shows you a quiet day, because the cloud work happened somewhere else. Open Settings > Usage when you want the cloud figures.
What makes a cloud session expensive?
The cost of any Claude Code session is the size of the context it sends multiplied by how many times it sends it. Cloud sessions tend to inflate both halves, for reasons that have nothing to do with the VM.
A long background run resends the whole conversation on every turn, and each tool call adds another request carrying that batch of results. A session left working for an hour therefore pays for its own history many times over. That is the same mechanism described in what actually consumes tokens in a Claude Code session, and it applies identically to local work. The difference is that nobody is watching a cloud session, so it runs much longer before someone stops it.
Re-reading the repository is the other big one. The VM clones the repository fresh and starts with no memory of whatever you explored yesterday. A vague instruction like "clean up the error handling" makes Claude search widely and read many files just to find where the error handling lives. A specific instruction like "add the retry wrapper from src/http/retry.ts to the three callers in src/sync/" sends a fraction of the tokens, because it removes the search.
/clear works in a cloud session outside a project's threads as it does locally: it starts a fresh conversation with empty context in the same cloud session. Clearing between unrelated tasks keeps the next task's context small. A thread you keep reusing without clearing carries everything that came before into every later request.
Idle gaps cost money too. A cloud session stops after a period of inactivity and its VM is reclaimed. Reopening the session from claude.ai/code provisions a fresh VM with your conversation history restored, though background work that was still running when the VM went away is not restored. Your first message after that gap also misses the prompt cache, which means the full context is reprocessed at full price instead of being read back cheaply. By default the main conversation's cache lasts one hour while you are inside your plan and five minutes once you are drawing on usage credits, but that is a default, not a rule: the promptCacheTtl setting or the CLAUDE_CODE_PROMPT_CACHE_TTL environment variable set to 1h keeps the hour under either billing mode, at a higher rate for cache writes. In a cloud session, environment variables go on the cloud environment. Subagents default to five minutes under both.
Controls that work inside a cloud session outside a project's threads: /context to see what is in the window, /clear to start over with an empty context, /compact keep the test output to summarize with a focus you choose, and /model sonnet or /effort with a level as its argument, so you stop paying Opus rates for mechanical work. Note that /compact reads the conversation it summarizes, so compacting a very large context is itself a large request. Subagents keep verbose output in a separate context window, but each subagent sends its own requests, so they shrink your main context rather than your total. The shape of that overhead is covered in what an agent harness adds to every request.
Running the same job locally
The local route for this work is the ordinary CLI, and the useful pattern is to split the job rather than pick one side.
Plan on your machine, where exploration is cheap to interrupt:
claude --permission-mode planIn plan mode Claude reads files and runs commands to explore, then proposes an approach without editing source. Commit the plan, push it, and hand the execution to a cloud session:
claude --cloud "Execute the migration plan in docs/migration-plan.md"One detail catches everyone once. The cloud VM clones your repository's GitHub remote at your current branch, not your local working copy, so anything you have not pushed does not exist as far as that session is concerned. Push first. The exception is a repository with no git remote, or a github.com repository the Claude GitHub App is not installed on: there Claude Code uploads a bundle of your local repository instead, including uncommitted changes to tracked files but not untracked files. On macOS, Linux, and WSL, the bundle leaves out uncommitted changes to files named like credentials or keys, including .env, *.tfvars, and *.pem, and to files managed by git filters such as Git LFS; native Windows uploads tracked changes as they are.
To finish a cloud session on your own machine, pull it down:
claude --teleportTeleport fetches and checks out the session's branch, then loads the conversation into your terminal; if the cloud session was compacted, it continues from the summary rather than copying the earlier messages. It requires a clean working tree and a checkout of the same repository, not a fork. After that the terminal has its own copy of the session: new work there stays local and does not appear back on claude.ai.
Running locally does not make a single token cheaper. What it buys you is control: /resume to come back later, no VM expiry, and the ability to press Escape the second Claude heads somewhere wrong. For work you will sit next to anyway, that control is usually worth more than the convenience. For a test suite that takes twenty minutes, or a task you want progressing while your laptop is shut, the cloud earns its place. Keeping the context small pays off in both places, which is the subject of managing the context window deliberately.
When changing plan is the real answer
Credits are the expensive way to buy capacity, because they are billed per token at standard rates while a subscription is a flat fee. If you top up every week, the arithmetic has already turned against you.
Two signals say the plan is the problem rather than your habits. The first is hitting the limit during ordinary work, before you have attempted anything unusual. The second is a credit balance that empties on the same schedule every month. Tighter prompts fix neither.
What to compare, and where the plans actually differ, belongs on the plan pages rather than in a paragraph here. Start with which plan matches the way you work, and if you are already on Max, the difference between Max and Ultra. Decide with your own figures from Settings > Usage open in front of you.
FAQ
Do cloud sessions cost extra on top of my Claude plan?
No. The cloud sessions documentation states that "there's no separate charge for the cloud VM" and that cloud sessions "share rate limits with all other Claude and Claude Code usage within your account". The machine is included. What you spend is tokens, drawn from the same pool your terminal and your chats use, unless a bonus credit sits in front of that pool and is spent first. The exceptions are work billed to usage credits, such as fast mode, ultrareview runs after the free ones, and Fable on Pro beyond its launch allowance, which cost extra wherever they run. Through November 16, 2026, Pro includes all Fable models up to 50% of weekly limits at no extra cost.
Why does my local /usage not show what my cloud session did?
Because its day and week figures are computed from session history on the machine you ran the command on. The documentation is explicit that "usage from other devices or claude.ai is not included", so cloud work never appears there. Open Settings > Usage on claude.ai, which is the authoritative view for both plan usage and credit spend.
Does my cloud session stop when the credit runs out?
It depends which credit. When the promotional cloud session credit ran out, or expired on 4 November 2026, cloud sessions kept running and went back to counting against your normal plan limits. Fable on Pro normally bills to usage credits, but from November 9 through November 16, 2026 the Fable 5.2 launch offer includes all Fable models on Pro up to 50% of weekly limits; usage credits apply after that allowance is used or the offer ends. When your plan limit is reached and usage credits are on and funded, work continues billed per token at standard API rates until your monthly spend limit, or until your balance runs out if auto-reload is off. When your plan limit is reached with no credits available, new work stops with a message such as "You've hit your session limit" that names the reset time. Anthropic documents the capped five-hour wrap-up allowance for eligible terminal, IDE, and Desktop Code sessions, not cloud sessions.
Is Claude Code cheaper to run locally than in the cloud?
Per token, no: the same request costs the same wherever the session executes, and the VM is free either way. In practice local sessions often use less, because you are usually watching them, you can stop a run that is going wrong, and no VM expiry forces a restart that reprocesses your history. A cloud session gets expensive when it is left to work unattended on a vague instruction.