---
title: Monitoring AI usage with AI Runtime Control
summary: >-
  Read the request and token usage that AI Runtime Control (ARC) records, in a
  summary view and a per-request logging view.
url: >-
  https://www.fastly.com/documentation/guides/platform/ai-runtime-control/monitoring-ai-usage-with-arc
---

> **IMPORTANT:** 
>
> This information is part of a beta release. For additional details, read our [product and feature lifecycle](https://docs.fastly.com/products/fastly-product-lifecycle#beta) descriptions.
>
>

AI Runtime Control (ARC) records each request it routes, so you can see how your applications and developers use AI over time. You can use ARC's usage views to:

- see your total request and token volume over a date range and how it trends over time.
- find which virtual keys and models account for the most usage.
- review the individual requests made by a given virtual key.

ARC presents this in two views: a summary view for usage at a glance, and a logging view for individual request records.

## Prerequisites

To monitor AI usage, you must be assigned the [superuser role](https://www.fastly.com/documentation/guides/organizations-and-accounts/account-members-and-permissions/about-user-roles-and-permissions/). Other account roles can't access ARC.

## Viewing the ARC summary

The summary view shows your request and token volume across a date range, with charts that break usage down by requests made and tokens consumed, as well as tables that summarize usage for virtual keys and LLM provider models. Use it to understand overall usage and to spot which keys and models drive it. You can filter what you see by model, provider, virtual key, and date range.

To view the summary, follow these steps.

1.   Log in to the [Fastly control panel](https://manage.fastly.com).

2.   Go to **Tools** > [**AI Runtime Control**](https://manage.fastly.com/ai-runtime-control/summary).

3. Click the **Summary** tab.

> **HINT:** To export a CSV copy of your AI usage in AI Runtime Control, click **Export** at the top of the Summary tab.

## Viewing the logs

The logging view lists the requests routed through ARC. For each request, you can open a session log and investigate usage in detail. Each record includes the virtual key, request count, models touched, input and output token counts, and a timestamp.

To view the logs:

1.   Log in to the [Fastly control panel](https://manage.fastly.com).

2.   Go to **Tools** > [**AI Runtime Control**](https://manage.fastly.com/ai-runtime-control/summary).

3. Click the **Logging** tab.
4. Search for a virtual key, or set the virtual key, provider, model, and date range filters to narrow the records.
5. Do one of the following:
   - To view the session log for a record, click the **view details** icon at the end of its row. The session log shows the full record as JSON: the message sent, the response returned, the session ID, and metadata that includes the model, virtual key, and input and output token counts.
   - To view the details and metrics for the virtual key that made a request, click the key's name in its row. For more information, check out our guide on how to view a virtual key's details.

> **HINT:** To save a copy of the session log, click **Copy** to copy the JSON or **Download JSON** to download it.