Skip to main content

NVIDIA Riva STT plugin guide

How to use the NVIDIA Riva STT plugin for LiveKit Agents.

Available in
Python

Overview

This plugin allows you to use NVIDIA Riva as an STT provider for your voice agents.

Quick reference

This section provides a quick reference for the NVIDIA Riva STT plugin. For more information, see Additional resources.

Installation

Install the plugin from PyPI:

uv add "livekit-agents[nvidia]~=1.2"

Authentication

The NVIDIA Riva plugin supports two authentication methods:

  1. NVIDIA API Key: Set NVIDIA_API_KEY in your .env file to use NVIDIA's cloud services.
  2. Self-Hosted NVIDIA Riva Server: Deploy your own NVIDIA Riva server and configure the plugin to communicate with it using the riva_uri parameter.

Usage

Use NVIDIA Riva STT in an AgentSession or as a standalone transcription service. For example, you can use this STT in the Voice AI quickstart.

from livekit.plugins import nvidia
session = AgentSession(
stt=nvidia.STT(
language="en-US",
model="parakeet-rnnt-1.1b",
),
# ... llm, tts, etc.
)

Parameters

This section describes some of the available parameters. See the plugin reference for a complete list of all available parameters.

languagestringOptional

BCP-47 language code for the speech recognition language. Common values include en-US, es-ES, fr-FR, de-DE, zh-CN, ja-JP, and more. See NVIDIA Riva documentation for a complete list of supported languages.

modelstringOptionalDefault: parakeet-rnnt-1.1b

The NVIDIA Riva ASR model to use. Popular models include parakeet-rnnt-1.1b for streaming and parakeet-ctc-1.1b for batch processing. See NVIDIA Riva documentation for available models.

riva_uristringOptional

The URI of your self-hosted NVIDIA Riva server. If not provided, the plugin uses NVIDIA's cloud services (requires NVIDIA_API_KEY).

Additional resources

The following resources provide more information about using NVIDIA Riva with LiveKit Agents.