# Read Faustine Chat logs and usage

The **Monitoring** group of the Faustine Chat configuration has two tabs. **Logs** shows what your users did with the chat. **Usage** shows the tokens the chat consumed and an estimate of what your model provider bills for them.

## Before you begin

- Faustine Chat is set up for your workspace. See [Set up Faustine Chat for your team](https://docs.faustine.ai/apps/set-up-faustine-chat/).

## Read a conversation

1. In the portal, go to [**Faustine Chat**](https://auth.faustine.ai/go?to=%2Fapps%2Ffaustine-chat) and select **Configuration**.
2. Under **Monitoring**, select [**Logs**](https://auth.faustine.ai/go?to=%2Fapps%2Ffaustine-chat%2Fconfiguration%3Ftab%3Dlogs).
3. Narrow the list with the search box, which matches the conversation's title, and with the filters.
4. To read the messages, select the conversation. **Back to conversations** returns to the list.

The numbers above the list count the conversations and errors, and give the error rate and the average latency.

| Filter | Options |
| --- | --- |
| **Chat user** | **All users**, or one chat user |
| **Status** | **All statuses**, **Successful**, **Errors only** |
| **Error code** | **All error codes**, or one code |
| **Time** | **Any time**, **Last hour**, **Last 6 hours**, **Last 24 hours**, **Last 7 days** |

## Find the errors

To see only the replies that failed, select **Errors** next to **Conversations**. The count of errors appears on the tab. Filter them by **Error code** and **Time**.

These are failures of the chat itself, such as a reply the model provider refused. A call that the gateway denied or that your API answered with an error is in the gateway's logs, under the chat user's name. See [Logs](https://docs.faustine.ai/agent-gateway/logs/).

## See what the chat cost

1. Under **Monitoring**, select [**Usage**](https://auth.faustine.ai/go?to=%2Fapps%2Ffaustine-chat%2Fconfiguration%3Ftab%3Dusage).
2. Choose the window: **7d**, **14d** or **30d**.

| Figure | What it is |
| --- | --- |
| **Estimated cost** | An estimate of what your model provider bills for the chat in the window, with the chat's share of the whole workspace. It shows "no rate" when Faustine has no price for the model. |
| **Tokens** | The tokens the chat consumed. |
| **Turns** | The number of assistant replies. |

The amounts are estimates and can differ from your provider's invoice. Your model provider bills you directly.

For the usage of the whole workspace, see [Usage](https://docs.faustine.ai/workspace/usage/).

## Next steps

- [Usage](https://docs.faustine.ai/workspace/usage/): the tokens and cost of every Faustine product in the workspace.
- [Logs](https://docs.faustine.ai/agent-gateway/logs/): each call the chat made through a gateway.
- [Add and manage chat users](https://docs.faustine.ai/apps/chat-users/): disable a user or end their sessions.