Workspace Settings
Language models, default model, thinking effort, and security.
The landing page of the workspace administration area holds settings that affect every member.
Workspace name
Shown read-only. Set when the workspace was created. Contact support to change it.
Language models
Controls which AI models are available to members in MainBranch Chat.
Enable or disable models
Toggle each model on or off. Disabled models don't appear in any member's model picker.
Default model
Pick the model that's pre-selected for every new chat. Members can still override per chat using the model picker.
Thinking effort
For reasoning-capable models, set the default level of reasoning effort:
- Low — Fastest, less deliberation.
- Medium — Balanced.
- High, Extra-high, Max — Progressively more reasoning at the cost of speed.
This applies as the default when a member uses a reasoning model. Members can override per chat.
Some models may still be used for other tasks like meeting summarization or smart form filling. The selection on this page applies only to chat.
Security
Workspace email domain
When enabled, anyone signing up with your company email domain joins the workspace automatically as a member — no invitation needed. Useful for fast-growing teams.
When disabled, every member needs an explicit invitation from an admin.
The domain is captured automatically from your account's email address.
Saving changes
Changes save when you click outside an input or toggle. There's no separate save button.
What's not here
The most important admin tasks have their own pages:
- Members — invite, manage roles.
- Authentication — SSO, 2FA enforcement.
- Apps — data source integrations.
- Assistant Configuration — custom instructions and chat-level toggles.
- Governance — usage analytics.
- Audit Logs — activity history.
- Data Export — downloadable activity archives.