Assistant CoreAssistant Core
Voice

Model and Voice Settings

How to choose models and voices for Voice Agent in Standard and Realtime modes.

Overview

Users can choose personal model and voice preferences in Web App > avatar > Settings > Voice. If a field is left as Default, Assistant Core uses the default model for the user's current plan.

Each plan has a different list of available models. See AI Models for the model catalog, voice lists, and default models.

The Language model (LLM) in Standard mode is used only for voice turns. It does not change the Chat model in the Web App header. See Chat Model Selection if you want to change the model for text chat.

Admins enable voice, recording policy, and Stream Response Text in the Admin Panel. See Enable Voice.

Standard Mode

Standard mode runs a chained pipeline: listen to speech → transcribe to text → generate the AI answer → read the answer aloud with TTS.

SettingRoleHow to choose
Voice activity detection (VAD) modelDetects when the user starts and stops speakingKeep Default unless you need to tune turn-ending behavior
Speech recognition (ASR) modelTranscribes speech to textChoose a stronger model if transcripts are often wrong, especially for Vietnamese, accents, or noisy environments
Language model (LLM)Generates the answerUse a fast/cost-effective model for normal Q&A; choose a stronger model for reasoning, tool use, or long questions
Text-to-speech (TTS) modelConverts the answer into spoken audioChoose based on voice quality, latency, language support, and credit cost
VoiceThe spoken persona for Standard modeSelect this after choosing the TTS model; available voices depend on that model
Speech speedHow fast TTS audio is spokenSlow it down for careful listening; speed it up for quick operational use
Silence durationHow long to wait before ending the user's turnIncrease it if users pause often; decrease it for faster responses

Realtime Mode

Realtime mode uses a native voice model to process live audio directly with very low turn latency. Because of that, Realtime has its own model and voice selection. Realtime does not use the Standard-mode TTS voice.

SettingRoleHow to choose
Realtime modelHandles native live voice conversationChoose a low-latency model for natural conversation; choose a stronger model when instruction following or tool use is more complex
Realtime voiceNative voice for the selected Realtime modelSelect this after choosing the Realtime model; changing models may reset the voice because each model has its own voice list

If the Realtime section does not appear, the current plan does not include Realtime models. See Plans & Pricing for plan availability.

Credit Cost

Voice usage consumes credits based on the model and usage:

TypeHow it is charged
ASRBy audio duration
TTSBy character count or audio duration, depending on provider
Realtime VoiceBy session duration

See Credits & Usage to understand how credits are calculated.

If you are not sure what to choose, keep Default. Defaults are selected for your current plan and are usually balanced for quality, speed, and credits.

On this page