Claude Account Bans, Suspensions and Appeals

A long post argues that repeated bans of paying users in mainland China stem not from IP or usage patterns but from the region being unsupported, making wrongful bans nearly costless and risk controls highly tolerant.

The author argues that Claude's repeated bans of Chinese users are not about abnormal behavior but about near-zero cost of wrongful bans in unsupported regions, with accounts flagged at signup and purged in batches.

A post mocks Anthropic's rule that abusing Claude gets you banned, criticizing what it frames as an expectation that users submit and worship the model's lectures.

A chart shows about 1.25 million Claude accounts banned on October 5, nearly 20 times the daily average, with leaked emails showing many hit the Supported Countries Policy.

A viral post claims Anthropic carried out a mass ban of Chinese users during China's National Day holiday, suspending 1.25 million accounts on October 5 alone.

A user who fears bans and has bought stolen accounts recommends a site selling Claude 20x subscriptions and API access, claiming they are legitimate and come with a warranty.

After another Claude ban, the author shares a workaround using Claude Code and Workbuddy with Claude API access through easyrouter, a relay run by Fu Sheng's team, and says the video itself was made with this setup.

A user complains that foreigners can use a GPT without capability downgrades and a Claude they don't have to worry about suddenly losing to a ban.

The author argues most Claude account bans come from proxy misconfiguration — direct fallback rules, no TUN mode, QUIC going direct, or a proxy that doesn't start at boot — all of which leak the real IP to Anthropic.

A user questions the figure of 10 million Claude bans in half a year, arguing the ban wave pushed many users to Codex and that OpenAI no longer leads on either user base or model quality.

A user who made a "beauty of chemistry" video with Opus 5.5 that blended harmonies says Claude is too strong to give up, but with a sixth account banned, the video may be their last.

Agreeing that Anthropic can detect disguised environments, the author shares a setup that runs Claude Code on a US-based Grok Bot cloud machine and controls it remotely through Tailscale to avoid bans.