Gnani AI adds expression controls to multilingual Timbre speech model
Gnani says Timbre v3.0 lets users direct pauses, whispers and other vocal expressions through text cues across ten supported languages.
Gnani AI has introduced Timbre v3.0, a text-to-speech model that uses cues embedded in a script to control vocal delivery, including pauses, whispers, laughter and changes in pace.
The company says the model supports nine Indian languages and Indian-accented English, with uses ranging from customer-service voice systems to narration, learning content and video voiceovers.
Its October 14 product blog says the update is live in the Gnani app. Planned Pipecat and LiveKit availability should not be described as already delivered integration.
Gnani has not supplied pricing, independently tested speech-quality results or version-specific customer adoption figures. Its performance claims remain company-reported.
