> ## Documentation Index
> Fetch the complete documentation index at: https://help.desk.min-ai.ar/llms.txt
> Use this file to discover all available pages before exploring further.

# Home and usage

> Your space at a glance in Home, and how many tokens your agents used and what they cost in Usage.

## Home

**Home** is your space's summary:

* How many conversations are **Queued**, **With agents**, **With the team** and **Unread**.
* **Messages, last 14 days**: from **Customers**, **Agents** and the **Team**, per day.
* **This month**: **AI conversations** (conversations an agent handled, one per day), **Turns**, **Handed off**, **Input tokens** (and how many came from cache) and the **Estimated cost**.
* Open **Cases** and **Closed (7 days)**.

If you haven't connected a channel yet, Home takes you there.

## Usage

**Usage** shows the tokens your agents used and what they cost, at each provider's published prices. You pay in your provider account; this is so you know how much and where.

* Pick **By day** or **By week**, the **Period** and an **Agent** (or **All agents**).
* At the top, the totals: **Cost**, **Turns**, **Input tokens**, **Output tokens**, **From cache** and the cost **Per turn**.
* The cost chart, with **Show table** to see the numbers.
* **By agent**, **By model** and the **Conversations that cost the most**.
* At the bottom, the **Prices per million tokens** table, in dollars.

Costs are **estimated** with the prices published as of the date shown; with OpenRouter, the cost it reports for each call is used. If a model has no price in the table, its tokens count but its cost isn't added, and the page says so.

In a conversation, **Customer and cases → AI usage** shows its turns, tokens and cost.

> [!TIP]
>
> To lower the cost: use the cheapest model for simple questions, **Effort** on Low, and have each action read only the **Fields it reads** that the agent needs.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.