@geekbb: sub2api converts subscriptions to API traffic, you don't support this anymore? Will it be flagged by anti-fraud systems? Okay, okay, okay
Summary
The sub2api tool converts subscriptions to API traffic, but it could be flagged by anti-fraud systems, sparking community skepticism.
View Cached Full Text
Cached at: 08/21/26, 09:23 PM
sub2api converts subscriptions into API traffic—this method is no longer supported and may be flagged by fraud detection systems?
I see.
Tibo (@thsottiaux): We’ve looked into several reports regarding varying usage limits for Codex. Any changes to these limits would be made transparently with community engagement.
What we did observe is that among the users affected, many were using sub2api—converting subscriptions into API
Similar Articles
@ai_super_niko: https://x.com/ai_super_niko/status/2077612167634030853
CLIProxyAPI is a 42k-star open-source tool that converts AI subscriptions like Claude and ChatGPT into local API endpoints via OAuth login. It allows tools that only support API keys (such as ByteDance Trae, Tencent WorkBuddy) to share the same subscription, solving the pain point of reusing subscriptions across multiple tools.
Wei-Shaw/sub2api
Sub2API is an open-source AI API gateway platform for distributing subscription quotas across AI models, with warnings about compliance and terms of service for use.
@cloudi66: https://x.com/cloudi66/status/2090294715665293335
Introduces an open-source plugin named dsh that converts web subscriptions like ChatGPT, Claude, and Grok into standard AI model APIs, allowing users to fully utilize their existing subscriptions without additional API fees, and includes features such as real-time search, image generation, and video generation.
@yaohui12138: Codex slashing quotas—this is something anyone who's done product work can spot at a glance. The essence is forcing hig…
作者指出 Codex 官方正在暗中削减订阅额度，迫使高频用户升级付费层级；当定价无法匹配真实需求时，第三方中间商（如 Agent space）以约五折价格转售额度，AI 工具正重演当年云服务、话费和视频会员的灰色转售套路。
@9hills: At first, I didn't really get what Anthropic was trying to do — why detect Claude Code using third-party providers? What's the big deal? Later I realized how clever Anthropic is: by modifying the date format in the system prompt and comparing it with their own logs, they can precisely identify proxy station IPs and accounts. As long as your local configuration points to a third-party URL but the request ultimately goes back to Anthropic, it must be a proxy station. So they can batch-ban those proxy stations.
Anthropic precisely identifies requests from Claude Code that use third-party providers (proxy stations) by modifying the date format in the system prompt and comparing it with their own logs, thereby batch-banning proxy station IPs and accounts. This technique requires no network upload activity and can even detect proxy usage within enterprise intranet environments.