> For the complete documentation index, see [llms.txt](https://cortex-docs.paloaltonetworks.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://cortex-docs.paloaltonetworks.com/cortex-agentix/configure-cortex-agentix/data-management/manage-compute-units.md). # Manage compute units Cortex AgentiX uses compute units (CU) for API queries. When running Cortex Query Language (XQL) queries on your data sources using APIs, each XQL query API consumes CU based on the timeframe, complexity, and number of API response results. **Compute units usage** Cortex AgentiX provides a free daily quota of compute units (CU) allocated according to your license size. Queries called without enough quota will fail. To expand your investigation capabilities, you can purchase additional CU by enabling the Compute Unit add-on. After purchasing the additional CU, you can enable the add-on by selecting **Settings → Cortex AgentiX License → Addons**, hovering over the **Extended Compute Units** tile and clicking **Enable**. The Compute Unit add-on provides an additional 1 compute unit per day for a year, in addition to your free annual quota. For example, if you have allocated 1,825 free annual CU, with the add-on, you will have a total of 2,190 annual compute units. The Compute Unit add-on is calculated on an annual basis, starting from the procurement of your add-on license. The minimum purchase amount is 50 compute units. You can configure the daily consumption limit for your compute units according to your organizational needs and change it when needed. For example, you can set a lower limit on a daily basis, and during an incident investigation, you can change it to a higher limit that enables you to consume more compute units. Your unused compute unit balance cannot be transferred from one licensing period to the next. ### View and manage your compute units usage You can view and manage your compute units usage at **Settings → Configurations → Data Management → Compute Unit Consumption**. {% hint style="info" %} ### Important Compute units consumption currently applies only to XQL queries. Agentic and LLM information is shown only for informational purposes and does not consume compute units. {% endhint %} #### Widgets The following widgets present information about your compute units consumption and enable you to manage the daily limit. {% hint style="info" %} ### NOTE Widgets include data for agents and LLM usage, but agents and LLM usage does not consume compute units at this time. Data is provided only for informational purposes. {% endhint %} The **Total annual usage** widget shows the number of free compute units per license year, the number of purchased compute units per license year, and the ratio of used compute units to your yearly total compute units. The **Daily limit** widget shows the day’s consumption of compute units. If you have **Edit** permissions for Public APIs, you can click **Set daily limit** to customize the daily limit to meet your organizational needs. The default daily limit is the annual quota divided evenly. * **Divide annual quota evenly**: Total annual compute units divided by 365. * **1% of annual quota**: 1% of the total annual compute units. * **No limit** * **Custom**: Configure a daily amount that is equal to or greater than your daily average calculated over a year (annual total/365). Use only integers. The **Consumption by category over time** widget, by default, displays the last 30 days. You can click individual categories under the graph to filter. You can also change the timeframe to the last 12 months by filtering by timestamp in the **Compute Units Usage** table below. The daily compute units are calculated at 00:00 UTC time. The red line represents your daily limit for that day. If you change the daily limit multiple times on a specific day, the displayed limit is the last number you configured on that day.\ ​\ The **Usage by type default**, by default, displays the last 30 days. You can click individual categories in the graph to filter. You can also change the timeframe to the last 12 months days by filtering by timestamp in the **Compute Units Usage** table below. #### Compute Units Table In the **Compute Units Usage** table, you can filter all the requests that were executed on your tenant. You can filter and sort according to the following fields:
* **ID**: Unique identifier representing the executed XQL API query or request. * **Timestamp:** Date and time of execution. * **Type**: Indicates the type of request. * **Trigger/PAPI Key ID**: * For API calls: PAPI Key. * For manual actions: User. * For automated actions: Rule or playbook. * **Query/Prompt**: The query or prompt. * **Compute Unit Usage**: How many units were used. * **Category:** XQL Queries, Agents & LLM. * **Trigger Type:** The type of source of the query or prompt. For example, automation rule or playbook. * **Billable:** Whether the query was deducted from your compute units. Requests that are non-billable are displayed for informational purposes and do not affect your daily limit or compute units balance. ### Investigate the XQL API results. In the **Compute Units Usage** table, locate an XQL API query, right-click, and select **Show results**. The query is displayed in the query field of the Query Builder, where you can view the query results. For more information, see [How to build XQL queries](/cortex-agentix/reference-and-developer-docs/cortex-agentix-xql/build-xql-queries/how-to-build-xql-queries.md). --- # Agent Instructions This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com. ## Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://cortex-docs.paloaltonetworks.com/cortex-agentix/configure-cortex-agentix/data-management/manage-compute-units.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.