Claude Code says Your account is on hold: the account is suspended, not your login
As of November 19, 2026, the Your account is on hold message in Claude Code does not mean your login expired: the Claude account behind that login has been suspended, and signing in again with the same account will not clear it. The official guidance is short: open https://claude.ai/restricted from the message to view the details or appeal, and if you have an API key that the hold does not affect, set it and keep working while the hold is resolved.
Updated 2026-11-19
One key for Claude, GPT and leading Chinese models, billed per token, up and running in minutes.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
Four points to sort out first
account_on_hold: the structured code in -p and the Agent SDK
In non-interactive mode (-p) and the Agent SDK, the structured error code for this message is account_on_hold. Match on that code in scripts, alerts, and logs instead of string-matching the human-readable text.
Older builds reported it as an expired login
Before v2.1.235, Claude Code reported a held account as Login expired · Please run /login, and those recovery steps cannot clear a hold, so users on older builds are the ones most likely to misread it as an authentication failure.
The revised Supported Regions Policy takes effect
Anthropic's Supported Regions Policy is marked Effective Nov 12, 2026 and treats physical presence in an unsupported region as unsupported use, even when the account itself sits in a supported country.
Signing in again with the same account
The official docs say the hold sits on the account rather than the login, so signing in again with the same account cannot clear it and /login is not the fix.
What Your account is on hold actually means
As of November 19, 2026, the message means the Claude account behind your current Claude Code login has been suspended by Anthropic. It is not a network fault, not an expired key, and not a broken login session. Claude Code shows it at two moments: when it tries to renew your saved login and learns of the hold, it prints Your account is on hold and can't use Claude Code.
View details or appeal: https://claude.ai/restricted; when a sign-in you complete in the browser reports the hold, it prints Your account is on hold and can't sign in to Claude Code. View details or appeal: https://claude.ai/restricted. Both variants end at the same destination, https://claude.ai/restricted. Getting this distinction right is what stops you from burning an afternoon on /login, a CLI reinstall, or a different terminal.
What changed in November 2026: regions policy, fixed error code, slower appeals
Three changes in the two weeks before November 19, 2026 bear directly on this error.
- 1Tighter regional scope: the Supported Regions Policy, marked Effective Nov 12, 2026, treats physical presence in an unsupported region as unsupported use, even for staff of otherwise supported entities; the Usage Policy update taking effect the same day likewise bars use of Claude from unsupported locations and by entities incorporated, headquartered, or majority-controlled there.
- 2Stable reporting: non-interactive mode (-p) and the Agent SDK return the structured code account_on_hold for the same state, and since v2.1.235 the message is no longer shown as Login expired.
- 3Slower appeals: the help center says replies are slower than usual right now because of a recent launch and a spike in email volume. In community threads, discussion clustered around 2026-11-12 (a count from a single community account, unverified), and some users say a verification prompt appeared right after a successful payment.
Three dates, from identity verification to the regions policy
Anthropic updated its identity verification help article: Persona is the verification partner, individuals need a physical government-issued photo ID and may be asked to take a live selfie with a phone, and the same page notes that an account can still be banned afterwards for repeated Usage Policy or Terms of Service violations, with the appeal form as the route if you believe it is a mistake.
The 2026 Usage Policy update was published. Its section on clarifying the Supported Regions Policy states that the policy prohibits use of Claude by people physically located in an unsupported region, by entities incorporated or headquartered in one, and by entities majority-owned or controlled from one, and the post announces that the updated policy takes effect on November 12.
The revised Supported Regions Policy (Effective Nov 12, 2026) and the updated Usage Policy both take effect. Community discussion clustered around 2026-11-12 (a count from a single community account, unverified), with users reporting Your account is on hold in Claude Code.
Confirmed in official sources vs community claims and open questions
Confirmed (Claude Code error docs, Anthropic help center, and policy pages)
Checkable as of November 19, 2026
- 1Meaning and wording. The docs state that the account behind your login is suspended, and list two messages, one covering a saved-login renewal and one covering a browser sign-in, both pointing at
https://claude.ai/restricted; a fresh sign-in does not help, since the restriction belongs to the account and not to the session. - 2Error code. In non-interactive mode (-p) and the Agent SDK the structured error code is account_on_hold, and before v2.1.235 the same state was reported as Login expired · Please run /login, whose recovery steps cannot clear a hold.
- 3Official what-to-do. Open the link in the message to view the hold's details or appeal it; if you have another Claude account or an API key that isn't affected by the hold, keep working while it is resolved, by running /login with that account or by setting the key with
ANTHROPIC_API_KEY. - 4Appeals and data. Officially listed ban reasons include creating an account from an unsupported location; you must be logged in to reach the appeal form; a Free, Pro, Max, or Ultra account banned for violating the Usage Policy can still log in to export data or delete itself, though the data it can access and export may be restricted depending on the violation on that account; if an organization you belong to is paused, click Request a review on that organization; for a warning you dispute, use the contact channel shown on the help center page.
- 5Credential priority. The help center puts it this way: Claude Code prioritizes environment variable API keys over authenticated subscriptions. A key set as an environment variable is billed at API pay-as-you-go rates even while a subscription is signed in.
- 6Regions policy. The policy page is marked Effective Nov 12, 2026 and covers physical location, entity registration or headquarters, and majority ownership or control.
Community claims and what Anthropic has not published
None of the following is an official conclusion, so do not act on it.
- 1Signing in again, switching terminals, reinstalling the CLI, or upgrading the build will not lift a hold: the restriction belongs to the account, not the login session, and a newer build only changes the message and the error code.
- 2A successful payment does not prove the account is fine: the identity verification page itself says an account may still be banned afterwards, and community users report a verification prompt appearing right after a completed payment.
- 3Appeal turnaround and success rates: the help center only says replies are slower than usual right now, with no timeline or ratio.
- 4The exact triggers, counting method, and manual review process behind a hold are not published.
- 5Do not merge the two errors: community posts often treat This organization has been disabled as the same hold, but that one comes from a stale
ANTHROPIC_API_KEYbelonging to a disabled Console organization, and the documented fix is to unset the environment variable.
Wait for the appeal vs keep shipping on a separate billing path
Wait for the appeal: no Claude Code and no claude.ai until the account is restored
Because the hold sits on the account, waiting covers both paths: the saved login cannot be renewed, a fresh browser sign-in fails, and every request from the local CLI bounces back to the same message. If this is your only account, delivery simply stops. Anthropic only says reply times are longer than normal right now, with no committed timeline, so betting a release date on a same-day restore is a bad plan.
Appeal through the official channel and keep working on an unaffected credential
The official what-to-do section already offers the second path: an API key that the hold does not affect lets you keep working while the hold is resolved, by setting the key with ANTHROPIC_API_KEY. The alternative is another account under your name that was never restricted, such as a separate work or organization account, reached with /login; the Usage Policy forbids creating new accounts to get around a ban.
A metered API key is a separate billing path from a subscription, so it keeps the code moving, while the account-side appeal, data export, and refund still follow Anthropic's own process.
Five steps: identify the error, appeal, then restore throughput
- 1Identify which message you got: a failure while renewing your saved login points to the renewal path, a failure during a browser sign-in points to the sign-in path; if you instead see the organization-disabled error, a stale key from a disabled Console organization is overriding your subscription, so unset
ANTHROPIC_API_KEYfirst. - 2Open the link from the message,
https://claude.ai/restricted, to read what this hold is about and which actions are offered. - 3To appeal, log in to the banned account in a browser and submit the form, since the form requires a logged-in session; use the same screen to export your data or delete the account yourself; if an organization rather than your own account is paused, click Request a review on that organization.
- 4To keep coding meanwhile, follow the official advice and switch credentials: set an API key that the hold does not affect, with
ANTHROPIC_API_KEY; with QCode, setANTHROPIC_BASE_URLtohttps://api.qcode.cc/apiandANTHROPIC_AUTH_TOKENto your cr_ key. - 5Finish with a check: run /status in Claude Code to confirm the active credential is the one you intend, and remove retired keys from your shell profile so the next session is not misread as an organization-disabled error.
Keep shipping while the appeal runs, on a metered API key
QCode is an independent API access service, not affiliated with Anthropic, and access is split by protocol.
- 1Claude models such as
claude-sonnet-5-5andclaude-fable-5-2go only to the Anthropic Messages endpointhttps://api.qcode.cc/api, the one Claude Code uses: setANTHROPIC_BASE_URLandANTHROPIC_AUTH_TOKENand start working. - 2The same cr_ key drives
gpt-6.2-solin Codex CLI, with base_url set tohttps://api.qcode.cc/openaiand wire_api = "responses". - 3
qwen3.8-maxandglm-5.3are served through the OpenAI-compatible chat/completions endpointhttps://api.qcode.cc/openai/v1, not through Codex's Responses API. Billing is per token, with per-model rates on /models. Not ready for a monthly plan? Pay-as-you-go Packs start at $4 for a one-time quota you use within 7 days; if you already have a Personal Plan, add a booster pack when you need more quota. To be explicit: QCode cannot lift, and does not affect, any action Anthropic takes on your Claude account, and the claude.ai appeal and any subscription refund stay inside Anthropic's own process and terms.
Frequently asked questions
Does signing in again fix the Your account is on hold error?
No. The docs put it plainly: Signing in again with the same account doesn't clear the message, because the hold is on the account rather than the login. Running /login, logging out and back in, or reinstalling the CLI cannot change that. The only thing that removes the message is the hold itself being lifted, and the entry point is the link in the error, https://claude.ai/restricted.
Is this a platform outage or has my account been suspended?
It is a suspension, not an outage. The official docs put it plainly: "The Claude account behind your login has been suspended." In the current release, only the two Your account is on hold messages carry the https://claude.ai/restricted link. Before v2.1.235, though, Claude Code reported a held account as Login expired · Please run /login, so upgrade Claude Code before you judge which error you have.
How long does an appeal take?
Anthropic has not published a turnaround time, so any specific number of days is a guess. The help center only states: "Our response times are currently longer than normal due to our recent launch and an increase in email volume." It asks for patience. Treat the wait as out of your control and keep your code moving on a separate credential.
Can I get a refund on a subscription after a ban?
Refunds follow Anthropic's own terms and subscription page. On data the official scope is narrower: a Free, Pro, Max, or Ultra account banned for violating the Usage Policy can still log in at claude.ai to export its data or delete itself, and you do not need to contact Anthropic, since those options sit on the same screen as the appeal form, but the data you can access and export may be restricted depending on the Usage Policy violation associated with the account.
Why does it fail when I travel to, or sit in, an unsupported region?
Because the policy treats being physically present in an unsupported region as unsupported use in itself. The Supported Regions Policy, Effective Nov 12, 2026, lists this as an unsupported use: "Use by persons while physically located in an unsupported region, including users and personnel of otherwise supported entities." So a company account in a supported country does not protect an employee who is travelling in an unsupported one. For the authoritative list of countries and regions, read the policy page.
If I switch to an API key while appealing, does that hurt my appeal?
The official what-to-do section itself suggests this route: if you have an API key that isn't affected by the hold, you can keep working while the hold is resolved by setting the key with ANTHROPIC_API_KEY. Anthropic has not said whether switching to an API key affects the outcome of an appeal, so the appeal is decided by Anthropic's own review. Watch the priority rule, though: an API key set as an environment variable is billed at API rates and overrides the authenticated subscription, so unset it once the work is done.
Sources
Claude Code official error documentation, code.claude.com/docs/en/errors (fetched 2026-11-19); Anthropic help center: bans, warnings, and appeals article 8241253 (support.claude.com, fetched 2026-11-19), Identity verification on Claude (support.claude.com, 2026-08-11), Manage API key environment variables in Claude Code (support.claude.com, 2026-05-05); Anthropic policy pages: Supported Regions Policy (anthropic.com/legal/supported-regions, Effective Nov 12, 2026) and the 2026 Usage Policy update (anthropic.com/news, 2026-10-08). Community material comes only from public discussion threads and is labeled as user reports in the text. Everything above was checked on 2026-11-19.
Appeal through the official channel, and keep shipping on a metered key
Create a QCode account for a cr_ key, set ANTHROPIC_BASE_URL=https://api.qcode.cc/api in Claude Code, and call multiple models billed per token, with rates on /models.
Related reading
Pay-as-you-go Packs: no monthly plan, valid 7 days
Buy a one-time quota valid for 7 days or add a booster pack to a Personal Plan; per-model rates are on /models.
Claude Code Account Ban Guide
Why Claude accounts get banned, the official appeal and refund channels, and per-token API access.
Claude Code Without a Claude Subscription: Login Paths
The login paths the official doc lists, the credit rules on the Console side, plus the two QCode environment variables and the one-click script.
Crypto payment guide
Nine coins, 56 networks, zero platform fee for AI coding API credits.
Anthropic's policy pages, help center, and Claude Code error documentation were checked on 2026-11-19 and are authoritative at anthropic.com, support.claude.com, and code.claude.com. QCode.cc is an independent API access service with no affiliation to and no endorsement from Anthropic or OpenAI, and it cannot lift or influence any action a vendor takes on your account. QCode model availability and live rates follow qcode.cc/models.