Microsoft.Extensions.AI Integration

Cross-SDK comparison

See the centralized MEAI documentation for feature matrices and comparisons across all tryAGI SDKs.

The OpenAI SDK implements IChatClient, IEmbeddingGenerator<string, Embedding<float>>, and ITextToSpeechClient from Microsoft.Extensions.AI.

Supported Interfaces

Interface	Description
IChatClient	Chat completions with text, streaming, tool calling, images, and structured output
IEmbeddingGenerator<string, Embedding<float>>	Text embeddings with custom dimensions
ITextToSpeechClient	Text-to-speech with GPT-4o mini TTS, built-in voices, custom voice IDs, and streaming audio

IChatClient

Basic Chat

using OpenAI;
using Microsoft.Extensions.AI;

using var client = new OpenAiClient("API_KEY");
IChatClient chatClient = client;

var response = await chatClient.GetResponseAsync(
    "Say hello!",
    new ChatOptions { ModelId = "gpt-4o-mini" });

Console.WriteLine(response.Messages[0].Text);

Streaming

await foreach (var update in chatClient.GetStreamingResponseAsync(
    "Count from 1 to 10.",
    new ChatOptions { ModelId = "gpt-4o-mini" }))
{
    Console.Write(string.Concat(update.Contents.OfType<TextContent>().Select(c => c.Text)));
}

System Messages

var messages = new List<ChatMessage>
{
    new(ChatRole.System, "You are a pirate. Always respond with 'Arrr!'."),
    new(ChatRole.User, "Hello!"),
};

var response = await chatClient.GetResponseAsync(
    messages,
    new ChatOptions { ModelId = "gpt-4o-mini" });

Tool Calling

var weatherTool = AIFunctionFactory.Create(
    (string city) => city switch
    {
        "Paris" => "22C, sunny",
        "London" => "15C, cloudy",
        _ => "Unknown",
    },
    name: "GetWeather",
    description: "Gets the current weather for a city");

var response = await chatClient.GetResponseAsync(
    "What's the weather in Paris?",
    new ChatOptions
    {
        ModelId = "gpt-4o-mini",
        Tools = [weatherTool],
    });

Structured Output (JSON Schema)

var schema = System.Text.Json.JsonDocument.Parse("""
{
    "type": "object",
    "properties": {
        "name": { "type": "string" },
        "age": { "type": "integer" }
    },
    "required": ["name", "age"],
    "additionalProperties": false
}
""").RootElement;

var response = await chatClient.GetResponseAsync(
    "Return info about Alice who is 30.",
    new ChatOptions
    {
        ModelId = "gpt-4o-mini",
        ResponseFormat = new ChatResponseFormatJson(schema, "person", "A person object"),
    });

Image Content

var message = new ChatMessage(ChatRole.User, []);
message.Contents.Add(new TextContent("What is in this image?"));
message.Contents.Add(new UriContent(
    new Uri("https://example.com/photo.png"),
    mediaType: "image/png"));

var response = await chatClient.GetResponseAsync(
    [message],
    new ChatOptions { ModelId = "gpt-4o-mini" });

Temperature, TopP, and Stop Sequences

var response = await chatClient.GetResponseAsync(
    "What is 2+2?",
    new ChatOptions
    {
        ModelId = "gpt-4o-mini",
        Temperature = 0f,
        TopP = 1f,
        StopSequences = ["5"],
    });

Additional Properties

Pass provider-specific parameters via AdditionalProperties:

var options = new ChatOptions
{
    ModelId = "gpt-4o-mini",
    Temperature = 0f,
};
options.AdditionalProperties ??= new AdditionalPropertiesDictionary();
options.AdditionalProperties["seed"] = 42;

var response = await chatClient.GetResponseAsync(messages, options);

IEmbeddingGenerator

Basic Embeddings

using OpenAI;
using Microsoft.Extensions.AI;

using var client = new OpenAiClient("API_KEY");
IEmbeddingGenerator<string, Embedding<float>> generator = client;

var result = await generator.GenerateAsync(
    ["Hello, world!"],
    new EmbeddingGenerationOptions { ModelId = "text-embedding-3-small" });

Console.WriteLine($"Dimensions: {result[0].Vector.Length}");

Batch Embeddings

var result = await generator.GenerateAsync(
    ["Hello", "How are you?", "Goodbye"],
    new EmbeddingGenerationOptions { ModelId = "text-embedding-3-small" });

Console.WriteLine($"Generated {result.Count} embeddings");

Custom Dimensions

var result = await generator.GenerateAsync(
    ["Hello, world!"],
    new EmbeddingGenerationOptions
    {
        ModelId = "text-embedding-3-small",
        Dimensions = 256,
    });

Console.WriteLine($"Dimensions: {result[0].Vector.Length}"); // 256

ITextToSpeechClient

using tryAGI.OpenAI;
using Microsoft.Extensions.AI;

using var client = new OpenAiClient("API_KEY");
ITextToSpeechClient ttsClient = client;

var response = await ttsClient.GetAudioAsync(
    "Today is a wonderful day to build something people love.",
    new TextToSpeechOptions
    {
        ModelId = "gpt-4o-mini-tts",
        VoiceId = "coral",
        AudioFormat = "mp3",
        Speed = 1.05f,
        AdditionalProperties = new()
        {
            [OpenAiTextToSpeechPropertyNames.Instructions] = "Speak in a cheerful and positive tone.",
        },
    });

var audio = response.Contents.OfType<DataContent>().Single().Data;

For low-latency playback, use streaming:

await foreach (var update in ttsClient.GetStreamingAudioAsync(
    "Stream this as audio chunks.",
    new TextToSpeechOptions
    {
        ModelId = "gpt-4o-mini-tts",
        VoiceId = "coral",
        AudioFormat = "pcm",
    }))
{
    foreach (var chunk in update.Contents.OfType<DataContent>())
    {
        var bytes = chunk.Data.ToArray();
        // Send bytes to your audio output pipeline.
    }
}

Custom Providers

All 15+ custom providers support the same MEAI interfaces:

using OpenAI;
using Microsoft.Extensions.AI;

// Any CustomProvider works as IChatClient and IEmbeddingGenerator
using var client = CustomProviders.Groq("API_KEY");
// or: CustomProviders.DeepInfra("API_KEY")
// or: CustomProviders.Together("API_KEY")
// or: CustomProviders.Azure("API_KEY", "ENDPOINT")
// or: CustomProviders.DeepSeek("API_KEY")
// or: any other CustomProvider

IChatClient chatClient = client;
var response = await chatClient.GetResponseAsync(
    "Hello!",
    new ChatOptions { ModelId = "llama-3.3-70b-versatile" });