Skip to content
Nappai Documentation

Vosk (local)

Vosk (local) is one of the components that act as the hearing of a voice agent: it turns what the caller says into text so the agent can understand it. It runs on the NappAI server: no credential, the audio is not sent to any outside provider and there is no per-minute cost. It understands Spanish and English, with less accuracy than Parakeet (local, multilingual) and the cloud engines. It transcribes while the caller speaks, so the word-based interruption settings of the Voice Agent, such as Minimum Words To Interrupt, work with it.

It plugs into the STT connector of the Voice Agent node. It holds the speech recognition settings the agent uses on every call.

When to use it

  • You want a local STT that transcribes while the caller speaks, so they can interrupt the agent with a few words.
  • Your calls are only in Spanish or English.
  • You are quickly testing an agent without setting up any outside account.

Configuration

FieldWhat it does
LanguageThe call’s language: es (default) or en.
Detect LanguageWhen on, the agent recognises whether the caller speaks Spanish or English and switches language with them. Off by default.

Advanced options

FieldWhat it does
Supported LanguagesLanguages detection can switch between. Only es and en are taken into account. Only used when Detect Language is on.
Model PathLeave it empty. Only needed if your installation’s technical team gives you another location for the speech model.

Output

The STT output holds the speech recognition settings. It connects to the STT connector of the Voice Agent node, or to Fallback STT when you turn on Fallback Engines and want it as a backup. The editor will not let you plug it into a TTS connector by mistake.

Example

A garage wants a simple agent that takes requests in Spanish.

  1. Add Vosk (local) and connect it to the STT connector of the Voice Agent.
  2. Leave Language as es.
  3. Test the agent in the playground by reporting a breakdown.

Good practices