Skip to main content

Choose the reading voice

When reading aloud is available, open Settings → Speech. These controls apply to task AI summaries; they do not change microphone capture or which AI provider transcribes a recording.

All screenshots
The example mission-control setup uses Male 3 at 1.25× for brisk Project Waddle briefings.
ScreenshotMobile view
Lotti Speech settings with Male 3 selected, the Supertonic 3 on-device model, and 1.25 reading speed

Choose a voice​

Use the Female / Male switch to browse the two groups, then select one of the five numbered voices in that group. Switching groups only changes the visible list; select a numbered voice to make it active.

All screenshots
Ten bundled voice styles are available: Female 1–5 and Male 1–5.
ScreenshotMobile view
Lotti Speech settings showing the five female on-device voice options

The selected voice is a device-local preference. That lets each device use a voice that suits its speakers and listening environment without changing the choice on your other devices.

Model download and privacy​

Lotti currently uses Supertonic 3 for reading aloud. Synthesis runs on the device rather than sending the summary to a cloud speech service. The model is large and is downloaded the first time it is needed, so allow time, network bandwidth, and storage for that initial setup.

Set the reading speed​

Choose a speed from 0.5× through 2×. A setting of 1× is the natural rate; lower values slow the briefing down and higher values shorten it. The selected speed is also stored only on the current device.

The read-aloud button appears only when the feature is enabled, the platform supports the speech engine, and the task has a non-empty AI summary.

For spoken input and saved recognition results, see Record audio.