How are AI credits managed?

Energy is required to power the complex reasoning of AI, and Procore uses credits to measure and manage that consumption. Learn how Procore is handling AI credits during the Limited Availability phase of Datagrid AI, which is subject to change.

How do I get credits?

Starter credits: Procore is providing each participating company with a starter pool of credits that refresh at the beginning of each month. Depending on the size of your project data, these credits may allow you to index a project and begin using agents.

Credit upgrades: If your account exhausts its credit limit, project-specific AI features are soon paused. Please contact your Procore account representative to review your options.

Which actions cost credits?

Datagrid AI consumes credits in relation to how hard it has to think about your request and how much of your project data it is ingesting, Learning, syncing, or considering when making decisions.

• Engagement: The AI consumes credits when users engage with it by asking questions, interacting via SMS, or generating new agent templates.

  • Accounts with Datagrid Enterprise may see these actions labeled in the metrics report as, Agent Interaction, Agent SMS Interaction, Knowledge Search, or AI Agent Template Generation.

  • Mode affects credits: Explore the three Datagrid AI modes below. Each time you send a question or request to Datagrid AI, you can select your preferred mode depending on whether you want to optimize for speed or accuracy of a complex task.

• AI Mode	Description	Credit cost
  Ask	Optimized for speed.	Lowest
  Extended	Enables deeper analysis of multiple data points for higher accuracy with nuanced requests.	Medium
  Execute	Plans and executes actions and multi-step tasks across data to assist with complex problems.	Highest
  Auto	Datagrid's reasoning engine dynamically selects a mode based on the complexity and context identified in your prompt.	Variable

  Note: Only 'Execute' and 'Auto' mode allow you to select from the Actions menu.

  Intelligence and data ingestion: The AI consumes credits when Learning (indexing) your data. This includes pulling data, processing those files, and the technical learning (vectorization) required to make that information searchable and ready for when a user engages with it. This process is construction-aware, ensuring your data is organized with an inherent understanding of industry relationships and logic.

  • Accounts with Datagrid Enterprise may see these actions labeled in the metrics report as, Run Import Pipeline, Dataset Processing, or Agent Learning (Vectorization).

  • Tip: Since Learning a Procore project for the first time may use a large portion of your monthly allotment, we recommend prioritizing your most critical project for initial Learning (indexing) to ensure you have enough credits remaining for daily AI interactions. See Enable Datagrid AI for more details. 

• Maintenance: A relatively small daily syncing cost associated with keeping your data hosted and fresh within the Datagrid engine so it's ready for retrieval without having to be re-processed every time you log in.

Where can I see credit usage?

Datagrid standalone product - Manage Plan modal

Admin of all tiers can click on the name of their plan tier (e.g., Pro or Enterprise) in the left panel to open the Manage Plan modal and see how many credits their teamspace has used versus how many remain.

Datagrid standalone product - Reports tab

In the Datagrid standalone app, Admin of Enterprise-level accounts can access the Reports tab to monitor credit consumption across the organization in the following ways:

• Overview visuals: High-level visualizations that group credit usage into functional categories, helping distinguish between background costs (like data ingestion) and active costs (like agent interactions).

• Historical trends: Utilize time-range filters (7, 30, 60, or 90 days) to track consumption patterns over time.

• Detailed metrics: View credit consumption by teamspace, users, and prompt activity. Visibility into user level consumption includes the action category (such as Dataset Processing or Run Import Pipeline), it's credit value, date, and from which teamspace it took place. Prompt activity includes the specific prompts and system tasks that triggered a credit cost.

Procore Company Admin tool

In the Company Admin tool, Company Admin can view an estimate of how many credits it takes to Learn (index) a particular project. 

Can I get basic AI support without using credits?

If you only need help navigating Procore (e.g., "How do I create a Submittal?"), you can save credits by using our AI-powered search in the global search bar at the top of the Procore application. Explore the differences between our three main AI services: