Visual Studio Code 1.131 has been released, coinciding with Microsoft Build 2026, and introduces built-in dictation, enhanced subagent visibility, and a hybrid Markdown editor. The new dictation feature uses Microsoft's offline Nemotron model for voice input across various platforms, eliminating the need for third-party extensions, while the subagent visibility improves user experience by allowing users to monitor subagent tasks directly in the parent conversation. Additionally, a new hybrid Markdown editor supports agent-actionable comments and inline editing, aiming to enhance workflows by merging documentation with executable instructions for AI agents. While this update marks significant improvements in voice-driven development and agent observability, some features remain experimental and are subject to platform restrictions
Visual Studio Code 1.131 Adds Built-In Dictation, Subagent Visibility, and Agent Polish
Visual Studio Code 1.131 drops on MS Build 2026 day, bringing built-in dictation, deeper subagent visibility, and a hybrid Markdown editor to the Stable channel. Native voice input now runs Microsoft's offline Nemotron model across chat, editors, and the terminal on Windows, Linux, and Apple Silicon Macs, removing the need for third-party extensions. Python Environments flips to default after hitting its rollout target, while agent workflows get better observability through a new parent-conversation overlay that tracks active subagents. Several features land behind experimental flags and platform restrictions, marking an incremental but meaningful shift toward voice-driven development and agent observability.
