Cursor Enterprise: Granular Model Controls & Smarter Spend Alerts

TL;DR

• Admins can now finely control AI model access, allowing or blocking specific models and providers.
• New soft spend limits provide flexibility with automatic alerts at usage thresholds.
• Usage analytics have been enhanced with user-level filtering and breakdown by product surface.
• Existing blocklists must be migrated by June 1st.

Cursor Enterprise is rolling out significant updates for administrators, aiming to provide more robust control over AI model usage and associated costs. The latest enhancements focus on granular model access, smarter spend management, and more detailed usage analytics. For teams heavily reliant on AI-assisted development, these features promise to streamline management and improve cost visibility.

A key update introduces a sophisticated new system for model access controls. Admins can now establish detailed allow or blocklists at both the model and provider level. This allows for blocking entire AI providers or even specific configurations of models based on parameters such as speed and context window size. Furthermore, enterprises have the option to automatically block new providers or model versions by default, ensuring that only approved technologies are used within the organization. It's important to note that customers currently using existing blocklists will need to migrate their configurations to the new system by June 1st. To access these controls, admins can navigate to their team model settings.

Beyond access controls, Cursor is also refining its spend management capabilities. Recognizing that hard limits can sometimes disrupt workflows, the platform now offers the option to set soft spend limits. This approach allows users to continue working while providing clear visibility into consumption patterns. Cursor will proactively monitor usage and send automatic alerts to users as they reach key thresholds, specifically at 50%, 80%, and 100% of their set limits, whether soft or hard. This feature aims to keep teams productive while empowering both users and administrators with an understanding of AI resource consumption. Setting up these new limits is accessible via the spend management settings.

The reporting capabilities have also seen a substantial upgrade with the introduction of more detailed usage analytics. Administrators can now filter usage data by specific users, providing a clear view of individual contribution and consumption. Additionally, the analytics can be broken down by various product surfaces, including clients, Cloud Agents, automations, Bugbot, and Security Review. This level of detail allows for more precise cost allocation, performance monitoring, and identification of areas for potential optimization. These enhanced analytics are available in the usage analytics tab within the Cursor dashboard.

Summary

• Cursor Enterprise admins gain granular control over AI model and provider access through new allow/blocklist features.
• Flexible soft spend limits replace rigid hard limits, accompanied by automated user alerts at 50%, 80%, and 100% thresholds.
• Enhanced usage analytics allow filtering by user and breakdown by product surface for better insights.
• Existing blocklists require migration to the new system before June 1st.

Source: Model controls, spend management, and usage analytics · Cursor