Cursor Design Mode adds multi-select and voice input for UI edits

By Harsh Desai15 June 2026

TL;DR

Cursor Design Mode now supports multi-selecting elements for group edits. Voice input allows continuous narration of UI changes.

What changed

Cursor Design Mode adds multi-select so agents map visual relationships and handle component groups in one pass. Voice input supports continuous narration of UI adjustments without pausing between runs. Vibe Builders, Basic Users and Developers apply these options straight in the browser workflow.

Why it matters

Vibe Builders and Developers finish grouped UI edits faster than with single-select flows in prior Cursor versions during dashboard layout tasks. Basic Users gain from uninterrupted voice narration that keeps momentum across several changes at once.

What to watch for

Compare results against single-select mode inside the same Cursor browser session. Open a test page with multiple UI elements, run a multi-select edit plus voice narration, then review the agent output for relationship accuracy.

Who this matters for

Vibe Builders: Multi-select UI elements to let agents grasp visual relationships and edit component groups.

Harsh’s take

Cursor is doubling down on the agentic IDE experience by removing the friction of sequential editing. Multi-select is a massive win for layout work: it provides the model with the necessary spatial context to understand how components relate to each other. This prevents the common issue where an agent fixes one button but breaks the alignment of the surrounding container.

The addition of continuous voice narration is equally important. It allows for a stream of consciousness workflow that matches how designers actually think. By decoupling the input from the execution cycle, Cursor is moving toward a more fluid, conversational interface for front-end development.

This is a clear signal that the future of UI work is about high-level intent rather than manual property tweaking.

by Harsh Desai

Source:cursor.com