🛠️ How-to guide

<aside> 🧠

WHAT IT DOES

The Models area lists the AI models your agents can use — Zammo's built-in platform defaults plus any of your own Azure OpenAI deployments — and lets you choose the workspace default and track usage.

</aside>

Before you start

Anyone in the workspace can view models and usage. Adding models or changing the default needs a role that can edit knowledge — Editor, Content Manager, Publisher, or Admin (see Roles & Permissions).

What you'll see

• Platform defaults — Zammo's shared, read-only models. These power your agents out of the box, so you don't have to configure anything to get started.
• Your custom models — any Azure OpenAI deployments you've added.

Filter by All, GPT, Embeddings, or Disabled, and see which model is the current default for each type.

The Models page — Zammo's read-only platform defaults plus your own Azure OpenAI deployments, with the default for each type marked, and View usage / Add model.

Choosing the default

Select Set default on a model — or adopt a platform default as your workspace default. Agents pick up the change on their next turn. An individual agent can still override this on its Agent Settings tab.

Adding your own model

Select Add model to connect an Azure OpenAI deployment — this keeps usage on your own subscription. Provide:

• Type — GPT or Embeddings.
• Name and an optional Description.
• Endpoint URL and Deployment name from Azure.
• API key (stored securely, never shown again).

Run Test to confirm it connects — saving is blocked until the test passes — then Create.

Tracking usage

View usage shows tokens for the period (with a trend and forecast), prompt/completion/call counts, a spend map by agent and activity, and breakdowns by model and top conversations. Filter by date range or agent.