Organize users and access
Overview
A workspace is the basic unit for organizing resources, permissions, and billing in QwenCloud. Each workspace has its own set of API keys, model access, and usage quotas.
Workspaces let you:
- Isolate resources between departments, teams, or projects
- Control which models can be called and set per-model rate limits
- Manage API keys independently per workspace
- Track usage and billing separately
Workspaces and regions
Workspaces are managed separately in each region. A workspace, its API keys, and its model permissions exist only in the region where the workspace was created. After you switch regions in the upper-right corner of the console, select or create a workspace in that region. Workspaces in other regions are not copied or deleted; switch back to see them. See Regions and endpoints.
Each region has its own default workspace, created automatically when you open that region for the first time.
Create a workspace
1
Select a region
Use the region selector in the upper-right corner of the console to select the region where you want the workspace.
2
Go to the Workspaces page
Go to Settings > Workspaces.
3
Create a workspace
Click Create workspace, enter a workspace name (up to 30 characters), and click Submit.
4
Choose the service deployment scope
In Germany (Frankfurt), China (Hong Kong), Japan (Tokyo), and US (Virginia), select a service deployment scope. Choose a scope with a specific geographic boundary if you have data compliance requirements, or Global for a larger inference resource pool. API keys from the workspace call models in that scope. Singapore and China North 2 (Beijing) need no selection.The scope does not change the Base URL, which contains only the region.
5
Grant model access (optional)
Select Authorize all existing and future models (model calling only) to grant the workspace access to every model, including models launched later. You can change this after creation.
Manage workspaces
From the Workspaces page, you can:
- Search: Find workspaces by name or workspace ID.
- Edit: Click Edit next to a workspace to open the editing panel.
- Switch: Use the Current Workspace dropdown to filter the view, or use the workspace switcher at the bottom of the sidebar to change your active workspace.
| Label | Meaning |
|---|---|
| Current | The active workspace. It determines the data shown on the API Keys and Analytics pages. |
| Default | The account's default workspace, created at sign-up. It cannot be deleted. |
Edit a workspace
Click Edit on any workspace to open the editing panel with two sections:
Basic info
- Workspace Name: Rename the workspace (max 30 characters).
- Total API Keys: Shows the current number of API keys and the maximum allowed (such as 0/20).
Model Permissions
Control which models can be called using API keys in this workspace.
If you selected Authorize all current and future models(model calls only) when creating the workspace, no additional authorization is needed. Otherwise:
- Click Edit Models to open the model selection dialog.
- Search by model name, or filter by Model Author or model type.
- Select the models to add. Use Select All Models on This Page to check every model on the current page.
- Set rate limits for each model (Calls / min for RPM and Tokens / min for TPM, when the model supports token-based metering), then click Select (N) to confirm.
- Click Submit on the panel to persist the changes.
The model selection dialog is paginated. Selections on one page may not carry over when you navigate to another page. Use search and filters to narrow results, or save after each page.
Switch workspaces
You can switch your active workspace in two ways:
- Workspace switcher: Click the workspace name at the bottom of the sidebar to open the switcher. Select the workspace you want to use.
- Workspaces page: Use the Current Workspace dropdown to filter and select a workspace.
API keys and workspaces
Each API key belongs to one workspace. When calling models through the API, the key determines which workspace's model access and rate limits apply.
- Default workspace keys can call all available models.
- Sub-workspace keys can only call models that have been granted to that workspace.
Limits
| Resource | Limit |
|---|---|
| Workspaces per account | 10, including the default workspace |
| API keys per workspace | 20 |
| Workspace name length | 30 characters |
FAQ
How do I get the workspace ID?
Go to the Workspaces page. The workspace ID is shown in the Workspace ID column of the table. The workspace switcher shows only workspace names, not IDs.
You can also find the ID in:
- Request logs: Open any request's Details, switch to JSON, and read the
workspace_idfield. - Try AI: Open the Try AI page; the
workspaceIdparameter in the URL is the current workspace ID.