New to Telerik UI for WinForms? Start a free 30-day trial

Getting Started with SpeechToTextButton

Updated on Feb 10, 2026

This article demonstrates how to add the RadSpeechToTextButton control to your desktop application and configure it for speech recognition functionality. At the end of this tutorial, you will have a fully functional speech-to-text button that captures voice input and handle the recognized text.

SpeechToTextButton Getting Started

Setup Steps

Adding RadSpeechToTextButton to Your Form

To start using the RadSpeechToTextButton control, just drag it from the Toolbox and drop it onto your form. Alternatively, you can add the control programmatically in your form's constructor or Load event handler.

Setting up a RadSpeechToTextButton

C#
 RadSpeechToTextButton speechToTextButton = new RadSpeechToTextButton();
VB.NET
Dim speechToTextButton As New RadSpeechToTextButton()

Install Microsoft.Web.WebView2 NuGet package

Before using the RadSpeechToTextButton control, ensure that the Microsoft.Web.WebView2 NuGet package (minimum version 1.0.3650.58) is installed in your project. When the button is first loaded, the WebView2 may prompt the user to grant microphone access and allow listening.

This package is required for the default speech recognizer functionality and can be downloaded from nuget.org.

Handle the SpeechRecognized event to get recognized text

The RadSpeechToTextButton provides the SpeechRecognized event, which is raised each time the speech recognizer successfully processes audio input and converts it to text. This event is the primary mechanism for retrieving the transcribed text from the control.

C#
private void RadSpeechToTextButton1_SpeechRecognized(object sender, Telerik.SpeechRecognizer.SpeechRecognizerSpeechRecognizedEventArgs e)
{
    string fullText = e.FullText;
}
VB.NET
Private Sub RadSpeechToTextButton1_SpeechRecognized(sender As Object, e As Telerik.SpeechRecognizer.SpeechRecognizerSpeechRecognizedEventArgs)
    Dim fullText As String = e.FullText
End Sub

That's it! You now have access to the text that the RadSpeechToTextButton listens to.

Processing Recognized Text

Once the text is recognized in the SpeechRecognized event, you can display it in controls such as RadLabel, RadTextBox, and more.

More information about additional events available in the RadSpeechToTextButton, see the Events article.

The SpeechRecognizerSpeechRecognizedEventArgs class provides the following detailed information about the recognized speech:

FullText: Gets complete, accumulated, recognized text from the current listening session.
FullTextConfidenceScore: Gets the confidence score for the recognized text. A value between 0 and 1 that indicates the confidence level of the transcription. Higher values represent greater confidence. A value of -1 indicates that no confidence score is available.

Using the SpeechRecognized event

C#
 private void RadSpeechToTextButton1_SpeechRecognized(object sender, Telerik.SpeechRecognizer.SpeechRecognizerSpeechRecognizedEventArgs e)
 {
     if (e.FullTextConfidenceScore >= 0.5)
     {
         // High confidence - process the text
         this.radLabel1.Text = e.FullText;
     }
     else if (e.FullTextConfidenceScore < 0.5)
     {
         // Low confidence - notify user
         this.radLabel1.Text = "Low confidence: " + e.FullText;
     }
 }
VB.NET
Private Sub RadSpeechToTextButton1_SpeechRecognized(sender As Object, e As Telerik.SpeechRecognizer.SpeechRecognizerSpeechRecognizedEventArgs)
    If e.FullTextConfidenceScore >= 0.5 Then
        ' High confidence - process the text
        Me.radLabel1.Text = e.FullText
    ElseIf e.FullTextConfidenceScore < 0.5 Then
        ' Low confidence - notify user
        Me.radLabel1.Text = "Low confidence: " + e.FullText
    End If
End Sub

Continuous vs. One-Time Recognition

The RadSpeechToTextButton supports two recognition modes: Continuous Recognition (Default) and One-Time Recognition. These modes allow you to control how the speech recognition process behaves.

The recognition modes are defined through the IsContinuousRecognition property, which indicates whether the speech recognizer operates in continuous recognition mode.

For more information about the internal states and lifecycle of the control, see the States article.

Continuous Recognition (Default)

By default, when you press the RadSpeechToTextButton, the control enters continuous listening mode. In this mode:

The recognizer continuously processes speech input.
The listening session remains active until you press the button again to stop.
This mode is ideal for scenarios where you need to capture extended speech or multiple sentences.

One-Time Recognition

To switch to one-time recognition mode, set the IsContinuousRecognition property to false. In this mode:

The recognizer automatically stops after detecting a single utterance or a brief pause in speech.
This mode is suitable for capturing short voice commands or text entries.

Use One-Time Recognition Mode

C#
this.radSpeechToTextButton1.IsContinuousRecognition = false;
VB.NET
Me.radSpeechToTextButton1.IsContinuousRecognition = False

Language Support

The service used by the default WebView2 speech recognizer supports multiple languages. The default language comes from the current culture of the system, but this can be changed via the LanguageTag property. The property works with the IETF language tags (e.g., "en-US", "de-DE", "fr-FR", etc.).

C#
 this.radSpeechToTextButton1.LanguageTag = "en-US";
VB.NET
Me.RadSpeechToTextButton1.LanguageTag = "en-US"