Fact · Dictation in your apps
On a supported desktop platform, VS Code downloads a speech recognition model on first use, and after that speech recognition works locally and offline.
Checked against its source on October 17, 2026
In the source's own words
“On first use on a supported desktop platform, VS Code downloads the default nemotron-3.5-asr-streaming-0.6b speech recognition model. After the download completes, speech recognition works locally and offline”
Voice support in VS Code (accessibility voice documentation), Visual Studio Code documentation (Microsoft), 2026, Use built-in dictation.
Open the original- What it measures
- The web version of VS Code does not support the on-device model and streams microphone audio to a cloud service instead.
- What it does not show
- The first download needs an internet connection. The docs list the supported desktop platforms in a separate passage, and they describe the model as the default, which may change.