> For clean Markdown of any page, append `.md` to the page URL.
> For a complete documentation index, see https://docs.sarvam.ai/llms.txt.
> For full documentation content in one file, see https://docs.sarvam.ai/llms-full.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.sarvam.ai/_mcp/server.

# Managing Your Voice Library

As your roster of cloned voices grows, a little organization up front keeps the library usable instead of cluttered.

## Naming and Organizing Cloned Voices

* Give cloned voices specific, descriptive names at creation time - "Course Narrator v2" is far easier to find later than a default label once your library has a dozen entries.
* Rename or delete a cloned voice directly from the **My voices** tab.
* Use the heart icon on any voice row to favorite it, so your most-used personas surface quickly without re-filtering the table every time.
* If you maintain multiple versions of the same clone (e.g. after a re-recording), version the name explicitly (`v1`, `v2`) rather than overwriting - this keeps older projects that already selected `v1` unaffected.

## Voice ID for the API

Each cloned voice has a **Voice ID** (`svc-…`) on its card under **My voices** (also on Voice Cloning home). Use the copy control, then pass that value as `voice_id` on `POST /voices/clone`.

`GET /voices` and `GET /voices/{voice_id}` return the same IDs for the same org — list/get response field `id` is the value you send as `voice_id` when synthesizing. Voices created in Studio and via `POST /voices/create` share this library. See [Using a Cloned Voice](/creative-cloning-use#use-the-same-voice-in-the-api) and [Manage saved voices](/api/api-guides-tutorials/voice-cloning/overview#manage-saved-voices).

## Keeping Projects Consistent

> **Note**
>
> A voice you select in the Voice Library carries over into Text to Speech and Dubbing automatically. Changing it here changes it everywhere it's referenced.

* Settle on a persona per project before generating multiple assets - retroactively swapping a voice mid-project means re-generating everything that used the old one.
* Periodically review unused cloned voices and remove ones you no longer need, especially if they were created for a one-off project.