Claude Account Bans, Suspensions and Appeals

A user vents that OpenAI models other than astra feel as frustrating as domestic ones, and adds that two of their Claude accounts were banned.

A Claude API reseller describes the relay-station business: building a Claude account pool cost seven banned accounts and nearly ¥5,000 in one night, and most relays just resell or dilute upstream APIs.

Banned from Claude Code again, the author explains how a first-time Kiro Pro+ upgrade nets 2000 credits for Opus 5.5 at near zero cost, plus a cancel-before-Nov-1 warning.

A user says they are running Claude with no precautions, unsure when a ban will hit, and will revisit ban-avoidance advice only after it happens.

In a reply about a misunderstanding, the author calls Claude's bans on Chinese users an open fact and says their own main Claude account was banned without refund.

After buying $250 and $125 Claude subscriptions on two Apple accounts, the user's accounts were banned and Apple rejected both refund requests, leaving them asking what to do next.

A user finally got charged for Claude after cycling through three virtual cards, only to wake up to a disabled account, with the lost project context hurting more than the money.

The same user replies to another account, recounting how they cycled through three virtual cards to pay for Claude and still had the account disabled after being charged.

After their Claude account was banned, a user ran codex in ~/.Claude to reconstruct that day's usage and draft a sincere appeal, and the account was restored the next day.

A user reports their Claude account was not banned, but the Google account used to log in was, with both Google accounts bought from an account seller getting banned.

A reply argues a user banned from Claude should have checked their network setup first, blaming the user rather than the reporter.

A user reports their Claude account was banned the day after topping up with a virtual card, wiping tuned project context, calling Anthropic's risk control brutal.