Voice Providers
Corrected 19 September 2026 against docs/user-guide/ (Part 3, section 3.3): Deepgram Aura and PlayHT are not
offered. The voices come from ElevenLabs, Sarvam, OpenAI and Qwen, plus your own cloned voice.
The “Voice” of your agent is powered by leading Text-to-Speech (TTS) engines. Movoice AI gives you access to multiple providers to ensure the most realistic experience in any language.
Supported Providers
ElevenLabs
- Best for: Ultra-realistic, emotive voices.
- Languages: 29+ languages supported.
- Key Feature: High-quality “Professional” voices that sound indistinguishable from humans.
Sarvam AI
- Best for: Indian languages (Hindi, Odia, Tamil, Telugu, etc.).
- Languages: Optimized for 10+ Indian regional languages.
- Key Feature: Understands Indian accents and colloquialisms better than global providers.
OpenAI and Qwen
- Best for: Further voice choices. Not every voice can speak every language; the app hides or warns about voices that cannot.
Your own cloned voice
- Best for: Sounding like you. Record a sample (phone: Edit > Voice > Clone your voice); ElevenLabs builds the voice by default.
Tuning the Voice
In the Agent Studio, you can tune several parameters:- Stability: Higher stability makes the voice consistent; lower makes it more expressive.
- Similarity Boost: Controls how closely the generated audio matches the original voice sample.
- Speed: Adjust the pace of speech (recommend 1.0x to 1.1x for natural conversation).
