Prepare to call an API¶
Choose the service you need, then use the links below to obtain credentials, a model, or workspace information before opening a specific API page. Genesis and AI Studio use different service addresses and authentication headers; never mix them in one request.
Call a Genesis model¶
List models and obtain the model ID to use from the response.
Open Genesis and choose text generation, embeddings, or reranking.
Genesis requests use Authorization: Bearer <credential> and do not need a workspace ID.
Manage AI Studio workspace resources¶
Create a personal access token on the Access credentials page.
List workspaces and obtain the target workspace ID from the response.
Open AI Studio and choose the workspace resource to manage.
AI Studio requests use X-API-Key; workspace-scoped actions also require X-Workspace-ID.
Pass the personal access token in X-API-Key when calling AI Studio, and use a single authentication method. If a Bearer request reports that the login session has expired, check the request headers first. Signing in again or creating another token does not change how the request sends it.
If the first request succeeds but later requests return ErrMultipleAuthCredentials, check whether the client automatically saves and resends cookies. Keep API-key clients separate from browser login sessions. See Troubleshoot failed calls for details.
Continue reading¶
For a complete Genesis walkthrough, see Call Genesis models.
To construct a request field by field, open the matching page in API Reference.