Documentation
Governance & model policy
As a tenant admin you decide how your assistant behaves within your plan — which features are on, whether your users choose their model or get one you've picked, and where your limits sit. These controls live in your admin console and take effect for your tenant only.
Features within your plan
Your plan sets a ceiling; within it, you turn features on or off for your tenant:
- Guest access — let anyone chat without signing in, or require sign-in.
- Human handoff — allow users to reach a person, and give your team the operator inbox. Off by default until you turn it on. See Human support & handoff.
- Tools — whether the assistant can take actions through your connected tools, and which. See Connecting your tools.
- Branding — your name, logo, colors, and welcome copy.
Think of it as two layers: your plan is the ceiling the platform sets; your settings are where you land inside it. You can move freely up to the ceiling, never past it.
Model policy: who chooses the model
You control which model answers your users, with two policies:
| Policy | What your users see | Use it when |
|---|---|---|
| User choice | A model picker in the composer; each user picks from the models you allow. | You want to offer a range and let power users decide. |
| Single default | One model, pinned. The assistant stays on it and won't switch. | You want consistent cost and behavior for everyone. |
You also set a model allowlist — the set of models available under either policy — so "user choice" still means your shortlist, never the entire catalog.
The policy is enforced end to end: under single default, a request to switch models is refused, so a user can't route around your choice. The picker only ever offers models on your allowlist.
Rolling out. Under a single-default policy the composer will show the pinned model as a static label with no picker at all — a small visual refinement. The policy itself is already enforced today; the picker just doesn't offer other models to switch to.
Limits
Your plan carries a usage limit. You can see how close you are on the Usage & analytics page, per provider and over time, so a busy month never surprises you.
Where these live
All of the above sit in your admin console — a governance panel for model policy and limits, a branding panel, a connectors panel, and a usage panel — each scoped to your tenant. Platform-wide settings (creating tenants, verifying domains) stay with the platform operator; your controls are your own tenant's.
Next
- Usage & analytics — track usage against your limit.
- The assistant experience — what your policy looks like to a user.
- Managing from chat — the same controls, conversationally.