Sign in to continue.
New here? Create an account, then link your own ChatGPT — or ask the owner for access to the shared one.
Your ChatGPT accounts, exposed as one OpenAI-compatible API.
Requests from your keys run on your own ChatGPT quota. Add a second account and the gateway moves to it by itself when the first one runs out. Nothing here touches anyone else's account.
Link several accounts to keep the gateway serving when one hits its plan limit. Failover uses the first healthy account and only moves on when it is exhausted; round robin and least used spread the load instead.
Pinning a model forces every request from that key to use it — the client's model field is ignored. Choose Auto to let the calling app decide.
| Name | Key | Model | Owner | Reqs |
|---|
| User | Role | Own ChatGPT | Uses mine | Status |
|---|
Point any OpenAI-compatible client here:
Models: