---

AssemblyAI C# .NET SDK

The AssemblyAI C# SDK provides an easy-to-use interface for interacting with the AssemblyAI API from .NET, which supports async and real-time transcription, as well as the latest audio intelligence and LeMUR models. The C# SDK is compatible with .NET 6.0 and up, .NET Framework 4.6.2 and up, and .NET Standard 2.0.

Documentation

Visit the AssemblyAI documentation for step-by-step instructions and a lot more details about our AI models and API. Explore the SDK API reference for more details on the SDK types, functions, and classes.

Quickstart

You can find the AssemblyAI C# SDK on NuGet. Add the latest version using the .NET CLI:

dotnet add package AssemblyAI

Then, create an AssemblyAIClient with your API key:

using AssemblyAI;

var client = new AssemblyAIClient(Environment.GetEnvironmentVariable("ASSEMBLYAI_API_KEY")!);

You can now use the client object to interact with the AssemblyAI API.

Add the AssemblyAIClient to the dependency injection container

The AssemblyAI SDK has built-in support for default .NET dependency injection container. Add the AssemblyAIClient to the service collection like this:

using AssemblyAI;

// build your services
services.AddAssemblyAIClient();

By default, the AssemblyAIClient loads it configuration from the AssemblyAI section from the .NET configuration.

{
  "AssemblyAI": {
    "ApiKey": "YOUR_ASSEMBLYAI_API_KEY"
  }
}

You can also configure the AssemblyAIClient other ways using the AddAssemblyAIClient overloads.

using AssemblyAI;

// build your services
services.AddAssemblyAIClient(options =>
{
    options.ApiKey = Environment.GetEnvironmentVariable("ASSEMBLYAI_API_KEY")!;
});

Speech-To-Text

Transcribe audio and video files

Transcribe an audio file with a public URL

When you create a transcript, you can either pass in a URL to an audio file or upload a file directly.

using AssemblyAI;
using AssemblyAI.Transcripts;

var client = new AssemblyAIClient(Environment.GetEnvironmentVariable("ASSEMBLYAI_API_KEY")!);

// Transcribe file at remote URL
var transcript = await client.Transcripts.TranscribeAsync(new TranscriptParams
{
    AudioUrl = "https://assembly.ai/espn.m4a",
    LanguageCode = TranscriptLanguageCode.EnUs
});

// checks if transcript.Status == TranscriptStatus.Completed, throws an exception if not
transcript.EnsureStatusCompleted();

Console.WriteLine(transcript.Text);

TranscribeAsync queues a transcription job and polls it until the transcript.Status is completed or error.

If you don't want to wait until the transcript is ready, you can use submit:

transcript = await client.Transcripts.SubmitAsync(new TranscriptParams
{
    AudioUrl = "https://assembly.ai/espn.m4a",
    LanguageCode = TranscriptLanguageCode.EnUs
});

Transcribe a local audio file

When you create a transcript, you can either pass in a URL to an audio file or upload a file directly.

using AssemblyAI;
using AssemblyAI.Transcripts;

var client = new AssemblyAIClient(Environment.GetEnvironmentVariable("ASSEMBLYAI_API_KEY")!);

// Transcribe file using file info
var transcript = await client.Transcripts.TranscribeAsync(
    new FileInfo("./news.mp4"),
    new TranscriptOptionalParams
    {
        LanguageCode = TranscriptLanguageCode.EnUs
    }
);

// Transcribe file from stream
await using var stream = new FileStream("./news.mp4", FileMode.Open);
transcript = await client.Transcripts.TranscribeAsync(
    stream,
    new TranscriptOptionalParams
    {
        LanguageCode = TranscriptLanguageCode.EnUs
    }
);

transcribe queues a transcription job and polls it until the status is completed or error.

If you don't want to wait until the transcript is ready, you can use submit:

transcript = await client.Transcripts.SubmitAsync(
    new FileInfo("./news.mp4"),
    new TranscriptOptionalParams
    {
        LanguageCode = TranscriptLanguageCode.EnUs
    }
);

Enable additional AI models

You can extract even more insights from the audio by enabling any of our AI models using transcription options. For example, here's how to enable Speaker diarization model to detect who said what.

var transcript = await client.Transcripts.TranscribeAsync(new TranscriptParams
{
    AudioUrl = "https://assembly.ai/espn.m4a",
    SpeakerLabels = true
});

// checks if transcript.Status == TranscriptStatus.Completed, throws an exception if not
transcript.EnsureStatusCompleted();

foreach (var utterance in transcript.Utterances)
{
    Console.WriteLine($"Speaker {utterance.Speaker}: {utterance.Text}");
}

Get a transcript

This will return the transcript object in its current state. If the transcript is still processing, the Status field will be Queued or Processing. Once the transcript is complete, the Status field will be Completed.

var transcript = await client.Transcripts.GetAsync(transcript.Id);

If you created a transcript using .SubmitAsync(...), you can still poll until the transcript Status is Completed or Error using .WaitUntilReady(...):

transcript = await client.Transcripts.WaitUntilReady(
    transcript.Id,
    pollingInterval: TimeSpan.FromSeconds(1),
    pollingTimeout: TimeSpan.FromMinutes(10)
);

Get sentences and paragraphs

var sentences = await client.Transcripts.GetSentencesAsync(transcript.Id);
var paragraphs = await client.Transcripts.GetParagraphsAsync(transcript.Id);

Get subtitles

const int charsPerCaption = 32;
var srt = await client.Transcripts.GetSubtitlesAsync(transcript.Id, SubtitleFormat.Srt);
srt = await client.Transcripts.GetSubtitlesAsync(transcript.Id, SubtitleFormat.Srt, charsPerCaption: charsPerCaption);

var vtt = await client.Transcripts.GetSubtitlesAsync(transcript.Id, SubtitleFormat.Vtt);
vtt = await client.Transcripts.GetSubtitlesAsync(transcript.Id, SubtitleFormat.Vtt, charsPerCaption: charsPerCaption);

List transcripts

This will return a page of transcripts you created.

var page = await client.Transcripts.ListAsync();

You can also paginate over all pages.

var page = await client.Transcripts.ListAsync();
while(page.PageDetails.PrevUrl != null)
{
    page = await client.Transcripts.ListAsync(page.PageDetails.PrevUrl);
}

[!NOTE] To paginate over all pages, you need to use the page.PageDetails.PrevUrl because the transcripts are returned in descending order by creation date and time. The first page is are the most recent transcript, and each "previous" page are older transcripts.

Delete a transcript

var transcript = await client.Transcripts.DeleteAsync(transcript.Id);

Transcribe in real-time

Create the real-time transcriber.

using AssemblyAI;
using AssemblyAI.Realtime;

var client = new AssemblyAIClient(Environment.GetEnvironmentVariable("ASSEMBLYAI_API_KEY")!);
await using var transcriber = client.Realtime.CreateTranscriber();

You can also pass in the following options.

using AssemblyAI;
using AssemblyAI.Realtime;

await using var transcriber = client.Realtime.CreateTranscriber(new RealtimeTranscriberOptions
{
    // If ApiKey is null, the API key passed to `AssemblyAIClient` will be
    ApiKey: Environment.GetEnvironmentVariable("ASSEMBLYAI_API_KEY"),
    RealtimeUrl = "wss://localhost/override",
    SampleRate = 16_000,
    WordBoost = new[] { "foo", "bar" }
});

Warning

Storing your API key in client-facing applications exposes your API key. Generate a temporary auth token on the server and pass it to your client. Server code:

var token = await client.Realtime.CreateTemporaryTokenAsync(expiresIn: 60);
// TODO: return token to client

Client code:

using AssemblyAI;
using AssemblyAI.Realtime;

var token = await GetToken();
await using var transcriber = new RealtimeTranscriber {
   Token = token.Token
};

You can configure the following events.

transcriber.SessionBegins.Subscribe(
    message => Console.WriteLine(
        $"""
         Session begins:
         - Session ID: {message.SessionId}
         - Expires at: {message.ExpiresAt}
         """)
);

transcriber.PartialTranscriptReceived.Subscribe(
    transcript => Console.WriteLine("Partial transcript: {0}", transcript.Text)
);

transcriber.FinalTranscriptReceived.Subscribe(
    transcript => Console.WriteLine("Final transcript: {0}", transcript.Text)
);

transcriber.TranscriptReceived.Subscribe(
    transcript => Console.WriteLine("Transcript: {0}", transcript.Match(
        partialTranscript => partialTranscript.Text,
        finalTranscript => finalTranscript.Text
    ))
);

transcriber.ErrorReceived.Subscribe(
    error => Console.WriteLine("Error: {0}", error)
);
transcriber.Closed.Subscribe(
    closeEvt => Console.WriteLine("Closed: {0} - {1}", closeEvt.Code, closeEvt.Reason)
);

After configuring your events, connect to the server.

await transcriber.ConnectAsync();

Send audio data via chunks.

// Pseudo code for getting audio
GetAudio(audioChunk => {
    transcriber.SendAudio(audioChunk);
});

Close the connection when you're finished.

await transcriber.CloseAsync();

Apply LLMs to your audio with LeMUR

Call LeMUR endpoints to apply LLMs to your transcript.

Prompt your audio with LeMUR

var response = await client.Lemur.TaskAsync(new LemurTaskParams
{
    TranscriptIds = ["0d295578-8c75-421a-885a-2c487f188927"],
    Prompt = "Write a haiku about this conversation.",
});

Summarize with LeMUR

var response = await client.Lemur.SummaryAsync(new LemurSummaryParams
{
    TranscriptIds = ["0d295578-8c75-421a-885a-2c487f188927"],
    AnswerFormat = "one sentence",
    Context = new Dictionary<string, object?>
    {
        ["Speaker"] = new[] { "Alex", "Bob" }
    }
});

Ask questions

var response = await client.Lemur.QuestionAnswerAsync(new LemurQuestionAnswerParams
{
    TranscriptIds =  ["0d295578-8c75-421a-885a-2c487f188927"],
    Questions = [
        new LemurQuestion
        {
            Question = "What are they discussing?",
            AnswerFormat = "text"
        }
    ]
});

Generate action items

var response = await client.Lemur.ActionItemsAsync(new LemurActionItemsParams
{
    TranscriptIds = ["0d295578-8c75-421a-885a-2c487f188927"]
});

Delete LeMUR request

var response = await client.Lemur.PurgeRequestDataAsync(lemurResponse.RequestId);