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

# Usage analytics

> Chart request, token, and error metrics for a project's inference traffic, split into serverless and dedicated usage.

The **Analytics** dashboard in the Together AI console charts inference requests, tokens, and errors over time for your selected project. Tabs split the traffic by usage type, so you can review serverless and dedicated endpoint usage separately. To track spend instead, use [cost analytics](https://docs.together.ai/docs/billing-cost-analytics.md).

[Analytics](https://api.together.ai/analytics)

## Choose a usage type

Select a tab to choose which traffic the dashboard charts:

* **All:** Every usage type in the project. The dashboard opens on this tab, which shows a total for each usage metric over the time range and a **Request volume & errors** chart.
* **Serverless:** Requests to [serverless models](https://docs.together.ai/docs/serverless/models.md).
* **Dedicated:** Requests to [dedicated endpoints](https://docs.together.ai/docs/dedicated-endpoints/overview.md), both on-demand and [reserved](https://docs.together.ai/docs/dedicated-endpoints/pricing.md#on-demand-vs-reserved).
* **Other:** Any other usage type in the project. This tab appears only when the project has that traffic.

On the **Serverless**, **Dedicated**, and **Other** tabs, select **Usage** or **Errors** to choose which charts appear.

## Group and filter the charts

Use **Group by** to split each chart into one series per value, or choose **None** to chart a single series. Select **Filter** to include only specific values. The options depend on the tab:

| Tab | Group by | Filters |
| - | - | - |
| **All** | Not available | Endpoint, API key |
| **Serverless** | Endpoint (default), API key | Endpoint, API key |
| **Dedicated** | Endpoint (default), API key, deployment | Usage type, endpoint, API key, deployment |
| **Other** | Endpoint, API key | Endpoint, API key |

On the **Dedicated** tab, the **Usage type** filter separates on-demand endpoints (**Dedicated**) from reserved endpoints (**Dedicated monthly**).

## Set the time range

The dashboard opens on the last 24 hours. Choose a unit to set how far back you can look and the size of each chart's time buckets:

* **Minutes:** Up to the last 60 minutes.
* **Hours:** Up to the last 24 hours.
* **Days:** A **Start** and **End** date, up to 30 days back.

## Next steps

> [card] [Cost analytics](https://docs.together.ai/docs/billing-cost-analytics.md)
>
> Track daily spend and billable units by product, project, or API key.

> [card] [Endpoint metrics and dashboards](https://docs.together.ai/docs/dedicated-endpoints/monitoring.md)
>
> Monitor a single dedicated endpoint and its deployments.


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