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

# Usage

> See the AI token use and the estimated cost for your whole company

The **Usage** screen shows how much your company uses the AI models. It shows the totals for the company, the use by each user, and the use by each provider.

Open the **Admin** group in the sidebar and click **Usage**. Only admins see this screen.

## Select the period

Click **7 days**, **30 days**, or **90 days** at the top of the screen. The screen shows **30 days** when it opens.

Click **Refresh** to load the newest figures. The time of the last load shows next to the button, for example **Updated 2:45:10 PM**.

## Company totals

Four figures show the totals for the period:

| Figure | Meaning |
| - | - |
| **Total tokens** | The tokens that all users sent to and received from the AI models |
| **Estimated cost** | The estimated cost of these tokens, in US dollars |
| **Active users** | The number of users who used the AI in the period |
| **Requests** | The number of requests to the AI models |

> [!NOTE]
>
> The cost is an estimate from RelayHub. Your provider invoice is the record of the cost.

## Usage by user

The **Usage by user** table shows each user with these columns:

* **User**: the name and the email
* **Tokens**: the tokens of the user
* **Share**: the percentage of all company tokens
* **Requests**: the number of requests
* **Cost**: the estimated cost

Use this table to see which users use RelayHub and how much. If no user used the AI in the period, the table shows "No usage recorded in this period".

## Usage by provider

The **Usage by provider** list shows each AI provider with its number of requests, its tokens, and its estimated cost. To manage the providers, see [Providers](https://docs.relayhub.ai/administration/providers/provider-hub.md).


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