Voxa.Audio.SileroVad 0.7.2-alpha

.NET 10.0 This package targets .NET 10.0. The package is compatible with this framework or higher.
This is a prerelease version of Voxa.Audio.SileroVad. 
dotnet add package Voxa.Audio.SileroVad --version 0.7.2-alpha
                    


                    

                
NuGet\Install-Package Voxa.Audio.SileroVad -Version 0.7.2-alpha
This command is intended to be used within the Package Manager Console in Visual Studio, as it uses the NuGet module's version of Install-Package. 
<PackageReference Include="Voxa.Audio.SileroVad" Version="0.7.2-alpha" />
For projects that support PackageReference, copy this XML node into the project file to reference the package. 
<PackageVersion Include="Voxa.Audio.SileroVad" Version="0.7.2-alpha" />
                    


                            Directory.Packages.props
                        

                    

                
<PackageReference Include="Voxa.Audio.SileroVad" />
                    


                            Project file
For projects that support Central Package Management (CPM), copy this XML node into the solution Directory.Packages.props file to version the package. 
paket add Voxa.Audio.SileroVad --version 0.7.2-alpha
                    


                    

                
#r "nuget: Voxa.Audio.SileroVad, 0.7.2-alpha"
#r directive can be used in F# Interactive and Polyglot Notebooks. Copy this into the interactive tool or source code of the script to reference the package. 
#:package Voxa.Audio.SileroVad@0.7.2-alpha
#:package directive can be used in C# file-based apps starting in .NET 10 preview 4. Copy this into a .cs file before any lines of code to reference the package. 
#addin nuget:?package=Voxa.Audio.SileroVad&version=0.7.2-alpha&prerelease
                    


                            Install as a Cake Addin
                        

                    

                
#tool nuget:?package=Voxa.Audio.SileroVad&version=0.7.2-alpha&prerelease
                    


                            Install as a Cake Tool

Voxa.Audio.SileroVad

ML-based voice activity detection for Voxa pipelines using the Silero VAD v5 ONNX model. Drop-in replacement for SilenceGateProcessor when energy-only filtering isn't enough — handles keyboard noise, fans, distant chatter, music in the background. Same emission contract as Voice Live: UserStartedSpeakingFrame / UserStoppedSpeakingFrame on speech transitions.

Install

dotnet add package Voxa.Audio.SileroVad --prerelease

The Silero VAD ONNX model (~2.3 MB, MIT-licensed) ships embedded in the assembly — no separate download or path config.

Quickstart

Swap SilenceGateProcessor for SileroVadProcessor:

using Voxa.Audio.SileroVad;
using Voxa.Speech;
using Voxa.Speech.Azure;

var pipeline = Pipeline.Build()
    .Source(new WebSocketAudioSource(ws, new WebSocketAudioOptions { InputSampleRate = 16000 }))
    .Then(new SileroVadProcessor())                            // ← replaces SilenceGateProcessor
    .Then(AzureSpeech.StreamingTranscription(azureSpeech))
    .Then(new MicrosoftAgentsProcessor(yourAgent))
    .Then(AzureSpeech.Synthesis(azureSpeech))
    .Sink(new WebSocketAudioSink(ws));

Tuning

new SileroVadProcessor(new SileroVadOptions
{
    SampleRate = 16000,             // also supports 8000
    ActivationThreshold = 0.5f,     // probability to open gate
    DeactivationThreshold = 0.35f,  // probability to close gate (hysteresis)
    MinSpeechWindows = 2,           // ~64 ms sustained → speech-start
    MinSilenceWindows = 8,          // ~256 ms sustained → speech-end
})

How it compares to SilenceGateProcessor

SilenceGateProcessor SileroVadProcessor
Pkg size ~10 KB ~2.3 MB model + ~50 MB ONNX runtime native binaries
Detects RMS amplitude Trained speech/non-speech classifier
Catches keyboard / fan / chair scrape no — high RMS, gets through yes — classified non-speech
Catches distant or quiet speech below RMS threshold no — gets dropped yes — classified speech
Threshold tuning per-room usually unneeded
Per-frame cost ~10 µs ~2–3 ms on CPU (still ~50× real-time at 16 kHz)
Cold start instant ~30 ms (embedded model load)

Use SilenceGateProcessor for the simplest case (clean office). Use SileroVadProcessor for production / noisy environments / mobile clients.

Constraints

  • Sample rate must be 16000 or 8000 Hz. Silero v5 doesn't support arbitrary rates. Audio at other rates is forwarded untouched with a warning — resample upstream if you need it.
  • Skip on the Voice Live composite path. Voice Live includes its own server-side VAD; running Silero on top is wasted work.

License

MIT. The bundled Silero VAD model is MIT-licensed by snakers4/silero-vad.

Product Compatible and additional computed target framework versions.
.NET net10.0 is compatible.  net10.0-android was computed.  net10.0-browser was computed.  net10.0-ios was computed.  net10.0-maccatalyst was computed.  net10.0-macos was computed.  net10.0-tvos was computed.  net10.0-windows was computed. 
Compatible target framework(s)
Included target framework(s) (in package)
Learn more about Target Frameworks and .NET Standard.

NuGet packages (1)

Showing the top 1 NuGet packages that depend on Voxa.Audio.SileroVad:

Package Downloads
Voxa

Batteries-included Voxa package. One package reference: AddVoxa(cfg) + MapVoxaVoice("/voice").UseDefaults() gives a working voice bot with STT, TTS, VAD, and an OpenAI chat agent — no knowledge of frames required.

GitHub repositories

This package is not used by any popular GitHub repositories.

Version Downloads Last Updated
0.7.2-alpha 101 7/10/2026
0.7.1-alpha 68 7/10/2026
0.7.0-alpha 71 7/10/2026
0.6.0-alpha 92 6/22/2026
0.5.0-alpha 71 6/13/2026
0.4.0-alpha 1,258 5/10/2026
0.3.0-alpha 73 5/9/2026