Using a Cloned Voice | Sarvam API Docs

Once processing completes, your cloned voice appears under the My voices tab in the Voice Library, and as a selectable persona in:

Text to Speech’s Cloned voices tab.
Dubbing’s Clone speaker voice strategy.

Everyone in your workspace sees your cloned voice under My voices and on the Instant Voice Cloning page. The Text to Speech Cloned voices tab lists only voices you created yourself.

Voice cloning works from a single ~10-second recording - one take is enough to generate a usable persona, though a cleaner, clearer reading generally produces more accurate results.

Where to Select It

In Text to Speech

Open the Cloned voices tab and select your voice.

In Dubbing

Choose Clone speaker voice as the speaker strategy, then select the cloned voice from the list.

See Managing Your Voice Library for removing clones you no longer need. Only the person who created a clone, an organization owner or admin, or a workspace admin can remove it.

Use the same voice in the API

Studio and the Instant Voice Cloning API share one voice library per workspace. Everyone in the workspace sees the same cloned voices, and a clone created in either place is available in both. Prefer the Python and TypeScript SDKs (voice_cloning.convert / voiceCloning.convert) for application code.

To call the API with a Studio-created voice:

Open Voice Library → My voices (or the Instant Voice Cloning home).
On the clone card, find Voice ID (format svc-…) and use the copy control.
Pass that value as voice_id on POST /voices/clone.

Voices you create with POST /voices/create also appear under My voices for everyone in the same workspace — the person who created one can select it in Text to Speech without recreating it in Studio.

See the Instant Voice Cloning API overview and Manage saved voices.