h0ax
Contributor
Contributor
After doing some research, I found that many cheap Claude API resellers seem to get their access through Chinese "transfer stations" rather than directly from Anthropic.
These services often use large pools of accounts, which may include hacked, stolen, or fraudulently created accounts. But that's only part of the story.
Some of these providers do offer models like Claude Opus and Fable, but many also mix in cheaper or even free models such as DeepSeek, Qwen, or GLM. In many cases, users have no way to know which model actually generated the response.
One investigation into this market found a complex system involving account farming, proxy networks, SMS verification services, and token-sharing platforms. This allows resellers to offer Claude access at a small fraction of the official price.
The biggest issue is quality and privacy. According to the CISPA Helmholtz study "Real Money, Fake Models", some relay services performed up to 47% worse than the official API. Researchers also found that some providers advertised premium Claude models while secretly using cheaper alternatives.
Privacy is another major concern. Because all prompts and responses pass through third-party servers before reaching Anthropic, they are logged by the relay provider. This means your conversations, prompts, uploaded data, and generated responses can be stored, analyzed, and potentially used for other purposes without your knowledge.
In short: if a Claude API provider is selling access for 10-20% of Anthropic's official price, there is usually a catch. The lower price often comes at the cost of reliability, response quality, privacy, and transparency.
So don't trust these services they provide here. They do exactly what I just explained
These services often use large pools of accounts, which may include hacked, stolen, or fraudulently created accounts. But that's only part of the story.
Some of these providers do offer models like Claude Opus and Fable, but many also mix in cheaper or even free models such as DeepSeek, Qwen, or GLM. In many cases, users have no way to know which model actually generated the response.
One investigation into this market found a complex system involving account farming, proxy networks, SMS verification services, and token-sharing platforms. This allows resellers to offer Claude access at a small fraction of the official price.
The biggest issue is quality and privacy. According to the CISPA Helmholtz study "Real Money, Fake Models", some relay services performed up to 47% worse than the official API. Researchers also found that some providers advertised premium Claude models while secretly using cheaper alternatives.
Privacy is another major concern. Because all prompts and responses pass through third-party servers before reaching Anthropic, they are logged by the relay provider. This means your conversations, prompts, uploaded data, and generated responses can be stored, analyzed, and potentially used for other purposes without your knowledge.
In short: if a Claude API provider is selling access for 10-20% of Anthropic's official price, there is usually a catch. The lower price often comes at the cost of reliability, response quality, privacy, and transparency.
So don't trust these services they provide here. They do exactly what I just explained


