​
Overview

Some PeachDesk Voice Studio engines come with a curated set of ready-made voices. Instead of cloning from your own audio, you pick a voice from a fixed catalog and the model speaks in that voice. No recording, no upload, no per-voice setup required. The models run on Bibha’s GPU infrastructure — you simply select an engine and a voice from a dropdown, and Voice Studio handles the rest. One engine offers built-in voices today: The other members of the VAANI family — VAANI, VAANI-ENGLISH, and VAANI-MULTILINGUAL — are cloning engines. They ship no predefined speaker table, so you give them a reference clip instead of picking from a catalog.
Looking to replicate a specific person’s voice instead? See Voice Cloning.

​
When to Use Built-in Voices

No reference audio

You don’t have (or don’t want to provide) a recording of the target voice.

Production reliability

Curated voices deliver predictable quality across any text input.

Speed

Skip the audio cleanup, sample preparation, and quality iteration loop.

Indian languages

VAANI-INDIC’s speakers are trained for Indic pronunciation and prosody, so you don’t have to source a native reference clip.

​
Creating a Built-in Voice Profile

Open Voice Studio from your PeachDesk dashboard, then head to the Voices tab to create a new voice profile.
1

Open the Voices tab

In Voice Studio, go to Voices and choose New voice.
2

Select 'Built-in voice'

When prompted for the voice source, pick Built-in voice instead of Clone from audio.
3

Choose the engine

Select VAANI-INDIC from the engine dropdown — it is the engine that ships built-in speakers.
4

Pick a voice

The voice catalog for the chosen engine appears, grouped by language. Preview each one by clicking it.
5

Name and save

Give the profile a name. No audio sample is needed — just save.
6

Generate

Use the profile like any other from the Generation tab.
Built-in voice profiles are tied to their source engine — the voice only exists for that model. When you switch engines, profiles from other engines are greyed out, and selecting one switches the engine back automatically.

​
VAANI-INDIC — 16 Voices Across 8 Indian Languages

VAANI-INDIC extends the multilingual backbone with an Indic adapter and ships two pre-extracted speaker conditionings — one female, one male — for each of the eight Indian languages it adds. Picking one involves no cloning: the speaker is baked into the model store, so the voice is identical on every generation. It runs on PeachDesk’s servers, so there’s nothing to install or download.
VAANI-INDIC also speaks English, inherited from the base model, but English ships no built-in speaker. For an English built-in-style voice, clone a reference clip with VAANI or VAANI-ENGLISH instead.

​
VAANI-INDIC at a Glance

A built-in speaker is a starting point, not a straitjacket. Use the Fine-tune panel on the Generation tab to adjust expressiveness and pacing, and the Effects tab to layer post-processing (including pitch and speed) onto a take. Each render is saved as a version, so you can compare deliveries side by side.

​
Cloning vs Built-in — Quick Decision

​
Limitations

Built-in voices are fixed — you can’t fine-tune or modify the underlying voice. If you want a voice that isn’t in the catalog, use a cloning engine and provide a reference sample under Clone from audio.
  • Built-in voices can’t be exported as standalone audio — a profile only stores metadata pointing to the same engine and voice ID.
  • The VAANI-INDIC catalog is set by the model’s shipped speaker conditionings — new voices appear only when PeachDesk rolls out a new model release.
  • VAANI, VAANI-ENGLISH, and VAANI-MULTILINGUAL have no built-in voices at all; querying their preset catalog returns an empty list by design.

​
Next Steps

Voice Cloning

Clone a specific voice from your own audio.

Generate Speech

Use a profile to generate audio on the Generation tab.

Build Stories

Compose multi-voice narratives in the Stories tab.