Team providers and models
What is it?
Team AI settings let Owners/Admins connect LLM providers with API keys and enable models the team can use across projects. This requires the team plan capability custom_models (paid plans).
Open from Team settings → AI providers (“Connect LLM providers and enable models”) → Manage AI.
Page framing: AI models: “Models enabled across your connected providers”
How to connect a provider
Open team AI settings
Team Settings → AI / Manage AI.
Connect a provider
Under Connect a provider, fill:
- Provider
- API key
- Base URL (optional)
Connect
Choose Connect provider.
Enable models
On the provider page:
- Enable / Disable models
- Disconnect the provider if needed
On the team models view, Enabled models can set reasoning to None, Low, Medium, or High, and remove models.
Permissions
Team AI management is part of team settings, so Owner/Admin access is required to open Team Settings. Connecting providers also requires the custom_models plan capability. Free plans use application presets only.
Expected behavior
- Without
custom_models, the connect-provider UI is unavailable; projects use presets. - Enabled team models become available when assigning models on a project.