Skip to content

Realtime Voice Model

Realtime Voice Model is the model of a Voice Agent Realtime, the one that hears, thinks and speaks at once, with any of NappAI’s providers: you choose Provider and the agent works exactly as with that provider’s own component, such as OpenAI Realtime, Gemini Live or Grok Voice.

The difference is that switching provider does not mean switching node: you change it in the field, and the node remembers the credential of every provider you already set up.

It plugs into the Realtime model connector of the Voice Agent Realtime node.

When to use it

  • You want to compare several realtime models with the same agent before settling on one.
  • You want to be able to move to another provider without touching the rest of the flow.

If the provider will not change, use its own component: it shows all its fields and its list of voices directly.

Configuration

FieldWhat it does
ProviderThe model’s provider: OpenAIRealtime (default), GeminiLive or GrokVoice. Choose it first: when you change it, the node asks for that provider’s credential.
Credential (required if the provider uses one)The credential of the chosen provider. The list only shows that provider’s credentials. See Credentials.
ModelThe provider’s model. When you switch provider it is filled with that provider’s default.
VoiceThe voice. Each provider has its own: when you switch provider it is filled with that provider’s default voice.
Turn EagernessHow soon the model takes its turn when the caller stops talking. Only with the providers that support it.

With GeminiLive, Language also appears: the language of the conversation. Empty: the model detects it.

Advanced options

FieldWhat it does
Provider SettingsOther fields of the provider’s own component, each with its own name. Fields the provider does not have are ignored.
Credentials by ProviderThe credential the node remembers for each provider. It fills itself in when you choose a credential.

Output

The Realtime model output holds the model settings. It plugs into the Realtime model connector of the Voice Agent Realtime node.

It does not work with the regular Voice Agent, which uses a separate model, hearing and voice. For that one there are Language Model, Speech to Text and Text to Speech.

Example

A customer service centre wants to compare OpenAI and Gemini with its realtime agent.

  1. Connect Realtime Voice Model to the Realtime model connector of the Voice Agent Realtime.
  2. Choose OpenAIRealtime and its credential, and test the agent in the playground.
  3. Switch to GeminiLive, choose its credential and a voice, and repeat the test.
  4. Keep the one that works best and deploy the flow.

Good practices

  • When you switch provider, check Voice: the previous provider’s voice does not exist in the new one.
  • The voice agent answers calls with the deployed version of the flow: after switching provider, deploy so that real calls use it.