Skip to content
Nappai Documentation

AssemblyAI STT

AssemblyAI STT is one of the components that act as a voice agent’s ears: it turns what the caller says into text so the agent can understand it. It uses AssemblyAI’s streaming speech recognition. It transcribes while the caller speaks and decides by itself when their turn is over, from the sound and from what they said, so the agent answers without waiting for a fixed silence.

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

When to use it

  • You want the agent to answer as soon as the caller stops talking, at a low cost.
  • The audio must be processed in the European Union: choose the eu region.
  • Your calls are in Spanish, English, French, German, Italian or Portuguese.

Configuration

You need an AssemblyAI API credential with your AssemblyAI account key. See Credentials.

FieldWhat it does
Credential (required)The AssemblyAI API credential used to connect to the service.
LanguageCode of the language the caller is expected to speak, for example es or en. Default es.
ModelThe AssemblyAI model. Default universal-streaming-multilingual, which understands Spanish, English, French, German, Italian and Portuguese. universal-3-6-pro is more accurate and costs more.
RegionWhere the audio is processed: us (default) or eu, inside the European Union.
Detect languageSwitches language mid-call if the caller does.

Output

The STT output holds the speech recognition settings. It plugs into the Voice Agent node’s STT connector, or into Fallback STT if you turn on Fallback Engines and want it as a backup. The editor won’t let you plug it into a TTS connector by mistake.

Example

A clinic wants its phone agent to answer quickly and the audio to stay in Europe.

  1. Add AssemblyAI STT and connect it to the Voice Agent’s STT connector.
  2. In Credential, choose your AssemblyAI API credential.
  3. Leave Language at es and set Region to eu.
  4. Try the agent in the playground.

Best practices

  • AssemblyAI bills usage to your own AssemblyAI account, for the length of the call. Check usage there.
  • The differences between local and cloud engines are explained in Quality, noise and latency.