Fact · Voice and AI
Microsoft says the fluid dictation experience in Windows Voice Access is powered by on-device small language models (SLMs).
Checked against its source on October 17, 2026
In the source's own words
“This experience is powered by on-device small language models (SLMs) to ensure fast and private processing.”
Fluid dictation (Voice Access, Windows help), Microsoft Support, 2026, Fluid dictation section.
Open the original- What it measures
- Microsoft's Windows help page for Voice Access, in the same section that describes what fluid dictation corrects.
- What it does not show
- Microsoft's own description of the architecture. The page does not name the models' sizes or their training data, and it describes on-device processing for this feature only.