Cursor Changelog

Recent highlights

Agent mode improvements

• Multi-file edits end-to-end. Cursor's agent can now plan, execute, and verify edits across multiple files in a single run.
• Custom model API keys. You can bring your own API keys for OpenAI, Anthropic, and other providers, giving more control over cost and model choice.
• Composer UI. A new planning and review interface for large changes, letting you approve or reject edits before they are applied.

Codebase understanding

• Improved indexing for large monorepos. Cursor now handles repositories with hundreds of thousands of files more gracefully.
• Better context suggestions based on recent edits and open files.

What this means for users

Cursor is becoming a stronger choice for teams working in large codebases who want a VS Code-native experience with agent capabilities. The Composer UI is particularly useful for reviewing big refactors before committing.

What to watch

Cursor releases frequently. The most impactful recent change is the ability to use your own API keys, which can significantly reduce cost for high-volume users.

Source: cursor.com/changelog