Cursor is an AI coding agent that plans, writes, and reviews code using AI agents that understand your entire codebase across Desktop, CLI, Web & Mobile.
This is the broad commercial model reported for Cursor. Product-level terms are listed below when available.
Trust profile
Cursor launched Cursor Router, an intelligent model routing system for enterprise and team plans. The feature automatically directs coding requests to the optimal AI model based on query task complexity, domain, and context. By dynamically assigning requests rather than using a single model, the company reports that customers can achieve frontier-level performance at significantly lower costs. Cursor Router is currently available across Cursor's desktop, web, iOS, CLI, and SDK platforms, with modes including Intelligence, Balance, and Cost to allow teams to manage their own cost-performance trade-offs.
Cursor has launched Cursor Router, an intelligent model classification system for Teams and Enterprise plans. The router analyzes incoming requests and dynamically routes them to the most appropriate AI model based on task complexity, context, and domain. It offers three optimization modes—Intelligence, Balance, and Cost—allowing users to balance performance versus token usage. Administrators can manage router availability, set default modes, and configure model allow/block lists per team or group. Testing indicates that the router can reduce costs by 30-50% while maintaining or improving task performance compared to using single-model configurations.
Cursor introduced functional improvements to its Slack integration, enabling agents to operate across multi-repo environments and interact with multiple Slack channels and threads. The update includes a new planning step where the agent provides a roadmap before commencing work, updated status reporting throughout task progression, and refined UI elements for better readability. These changes allow users to better manage complex, multi-repository tasks directly from within Slack.
Cursor released update 3.11, introducing 'Side Chats' which allow users to open secondary, persistent agent conversations alongside the main chat. The release also adds local search indexing for agent transcripts, allowing users to search through thousands of past conversations. The project and repository pickers were redesigned for improved scoping and integration with external platforms like GitHub and GitLab directly within the interface. Additionally, new hooks for Cloud Agents were added to support observability and control over agent behaviors including prompt, response, and subagent management.
Anysphere released a dedicated Cursor application for iOS. The app enables users to manage coding agents, implement code changes, investigate technical issues, and review pull requests from mobile devices. Features include voice-enabled interaction, visual feedback through image annotations, and remote monitoring of development tasks. The application supports integrations with Azure and Bitbucket for source control, facilitating software development and code review workflows outside of a standard desktop environment.