Package dev.langchain4j.model.azure

Class AzureOpenAiAudioTranscriptionModel

java.lang.Object

dev.langchain4j.model.azure.AzureOpenAiAudioTranscriptionModel

All Implemented Interfaces:

AudioTranscriptionModel

@Experimental
public class AzureOpenAiAudioTranscriptionModel
extends Object
implements AudioTranscriptionModel

Represents an Azure OpenAI audio transcription model, such as Whisper.

You can find a tutorial on using Azure OpenAI for speech to text at: https://learn.microsoft.com/azure/ai-services/openai/whisper-quickstart

Mandatory parameters for initialization are:

• endpoint: The Azure OpenAI endpoint URL
• authentication: Either apiKey, tokenCredential (Azure Active Directory), or an existing OpenAIClient
• deploymentName: The name of your Azure OpenAI audio model deployment

There are 3 authentication methods:

1. Azure OpenAI API Key: The most common method using an Azure OpenAI API key. Use the apiKey() method in the Builder.
2. Non-Azure OpenAI API Key: Use the OpenAI service directly (not Azure OpenAI). Use the nonAzureApiKey() method, which will automatically set the endpoint.
3. Microsoft Entra ID (Azure Active Directory): Authenticate using Azure credentials. Requires the com.azure:azure-identity dependency. Use the tokenCredential() method with an appropriate credential implementation like DefaultAzureCredential.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	AzureOpenAiAudioTranscriptionModel.Builder	Builder for AzureOpenAiAudioTranscriptionModel.

Constructor Summary

Constructors

Constructor	Description
AzureOpenAiAudioTranscriptionModel(com.azure.ai.openai.OpenAIClient client, String deploymentName, com.azure.ai.openai.models.AudioTranscriptionFormat responseFormat)	Creates a new AzureOpenAiAudioTranscriptionModel with the provided client and parameters.
AzureOpenAiAudioTranscriptionModel(AzureOpenAiAudioTranscriptionModel.Builder builder)	Creates a new AzureOpenAiAudioTranscriptionModel using the provided builder.

Method Summary

Modifier and Type	Method	Description
static AzureOpenAiAudioTranscriptionModel.Builder	builder()	Creates a new builder for AzureOpenAiAudioTranscriptionModel.
AudioTranscriptionResponse	transcribe(AudioTranscriptionRequest request)	Given an audio file, generates a transcription.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface dev.langchain4j.model.audio.AudioTranscriptionModel

transcribeToText

Constructor Details

AzureOpenAiAudioTranscriptionModel

public AzureOpenAiAudioTranscriptionModel(AzureOpenAiAudioTranscriptionModel.Builder builder)

Creates a new AzureOpenAiAudioTranscriptionModel using the provided builder. This is the recommended constructor for future compatibility.

Parameters:

builder - The builder containing all the configuration

AzureOpenAiAudioTranscriptionModel

public AzureOpenAiAudioTranscriptionModel(com.azure.ai.openai.OpenAIClient client,
 String deploymentName,
 com.azure.ai.openai.models.AudioTranscriptionFormat responseFormat)

Creates a new AzureOpenAiAudioTranscriptionModel with the provided client and parameters. Use the builder for more convenient construction.

Parameters:

client - The Azure OpenAI client

deploymentName - The deployment name of the audio model

responseFormat - The response format (can be null for default JSON format)

Method Details

transcribe

public AudioTranscriptionResponse transcribe(AudioTranscriptionRequest request)

Description copied from interface: AudioTranscriptionModel

Given an audio file, generates a transcription.

Specified by:

transcribe in interface AudioTranscriptionModel

Parameters:

request - The transcription request containing the audio file and optional parameters

Returns:

The generated transcription response

builder

public static AzureOpenAiAudioTranscriptionModel.Builder builder()

Creates a new builder for AzureOpenAiAudioTranscriptionModel.

Returns:

A new builder instance