從 Azure AI 推理 SDK 遷移到 OpenAI SDK

本文提供如何將您的應用程式從 Azure AI 推論 SDK 遷移到 OpenAI SDK 的指引。 OpenAI SDK 提供更廣泛的相容性、存取最新的 OpenAI 功能，並在 Azure OpenAI 和 Foundry 模型中統一簡化的程式碼模式。

備註

OpenAI SDK 是指連接到 openai的用戶端程式庫 （例如 Python openai 套件或 JavaScript npm 套件）。 這些 SDK 有自己的版本控制，與 API 版本分開 - 例如，Go OpenAI SDK 目前處於 v3，但它仍會連線到 URL 路徑中的 OpenAI v1 API 端點 /openai/v1/ 。

遷移的好處

遷移到 OpenAI SDK 具有以下幾個優勢：

• 更廣泛的模型支援：支援 Azure OpenAI 的 Foundry 模型以及 DeepSeek 和 Grok 等供應商所提供的其他 Foundry 模型
• Unified API：同時使用相同的 SDK 函式庫與用戶端，適用於 OpenAI 與 Azure OpenAI 端點
• 最新功能：可以取得最新的 OpenAI 功能，無需等待 Azure 的特定更新
• 簡化認證：內建支援 API 金鑰與 Microsoft Entra ID 認證
• 隱式 API 版本控制：v1 API 無需頻繁更新 api-version 參數

主要差異

下表顯示兩個 SDK 之間的主要差異：

層面	Azure AI Inference SDK	OpenAI 開發套件
用戶端類別	ChatCompletionsClient	OpenAI
端點格式	https://<resource>.services.ai.azure.com/models	https://<resource>.openai.azure.com/openai/v1/
API 版本	URL 或參數中的必要項目	不需要 （使用 v1 API）
模型參數	選用 （適用於多模型端點）	必填欄位：部署名稱
Authentication	僅限 Azure 認證資料	API key 或 Azure 認證憑據

設定

安裝 OpenAI SDK：

pip install openai

若要進行 Microsoft Entra ID 認證，請同時安裝：

pip install azure-identity

用戶端組態

OpenAI 開發套件

使用 API 金鑰驗證：

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.getenv("AZURE_OPENAI_API_KEY"),
    base_url="https://<resource>.openai.azure.com/openai/v1/",
)

Azure AI Inference SDK

import os
from azure.ai.inference import ChatCompletionsClient
from azure.core.credentials import AzureKeyCredential

client = ChatCompletionsClient(
    endpoint="https://<resource>.services.ai.azure.com/models",
    credential=AzureKeyCredential(os.environ["AZURE_INFERENCE_CREDENTIAL"]),
)

使用 Microsoft Entra ID 認證：

OpenAI 開發套件

from openai import OpenAI
from azure.identity import DefaultAzureCredential, get_bearer_token_provider

token_provider = get_bearer_token_provider(
    DefaultAzureCredential(), 
    "https://ai.azure.com/.default"
)

client = OpenAI(
    base_url="https://<resource>.openai.azure.com/openai/v1/",
    api_key=token_provider,
)

Azure AI Inference SDK

from azure.ai.inference import ChatCompletionsClient
from azure.identity import DefaultAzureCredential

client = ChatCompletionsClient(
    endpoint="https://<resource>.services.ai.azure.com/models",
    credential=DefaultAzureCredential(),
    credential_scopes=["https://cognitiveservices.azure.com/.default"],
)

聊天完成

OpenAI 開發套件

completion = client.chat.completions.create(
    model="DeepSeek-V3.1",  # Required: your deployment name
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "What is Azure AI?"}
    ]
)

print(completion.choices[0].message.content)

Azure AI Inference SDK

from azure.ai.inference.models import SystemMessage, UserMessage

response = client.complete(
    messages=[
        SystemMessage(content="You are a helpful assistant."),
        UserMessage(content="What is Azure AI?"),
    ],
    model="DeepSeek-V3.1"  # Optional for single-model endpoints
)

print(response.choices[0].message.content)

串流

OpenAI 開發套件

stream = client.chat.completions.create(
    model="DeepSeek-V3.1",
    messages=[
        {"role": "system", "content": "You are a helpful assistant."},
        {"role": "user", "content": "Write a poem about Azure."}
    ],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

Azure AI Inference SDK

from azure.ai.inference.models import SystemMessage, UserMessage

response = client.complete(
    stream=True,
    messages=[
        SystemMessage(content="You are a helpful assistant."),
        UserMessage(content="Write a poem about Azure."),
    ],
    model="DeepSeek-V3.1"
)

for update in response:
    if update.choices:
        print(update.choices[0].delta.content or "", end="")

嵌入技術

OpenAI 開發套件

from openai import OpenAI
from azure.identity import DefaultAzureCredential, get_bearer_token_provider

token_provider = get_bearer_token_provider(DefaultAzureCredential(), 
"https://ai.azure.com/.default")

client = OpenAI(
    base_url = "https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/",
    api_key = token_provider,
)

response = client.embeddings.create(
    input = "How do I use Python in VS Code?",
    model = "text-embedding-3-large" // Use the name of your deployment
)
print(response.data[0].embedding)

Azure AI Inference SDK

from azure.ai.inference import EmbeddingsClient
from azure.core.credentials import AzureKeyCredential

client = EmbeddingsClient(
    endpoint="https://<resource>.services.ai.azure.com/models",
    credential=AzureKeyCredential(os.environ["AZURE_INFERENCE_CREDENTIAL"]),
)

response = client.embed(
    input=["Your text string goes here"],
    model="text-embedding-3-small"
)

embedding = response.data[0].embedding

設定

安裝 OpenAI SDK：

dotnet add package OpenAI

若要進行 Microsoft Entra ID 認證，請同時安裝：

dotnet add package Azure.Identity

用戶端組態

使用 API 金鑰驗證：

OpenAI 開發套件

using OpenAI;
using OpenAI.Chat;
using System.ClientModel;

ChatClient client = new(
    model: "gpt-4o-mini", // Your deployment name
    credential: new ApiKeyCredential(Environment.GetEnvironmentVariable("AZURE_OPENAI_API_KEY")),
    options: new OpenAIClientOptions() { 
        Endpoint = new Uri("https://<resource>.openai.azure.com/openai/v1/")
    }
);

Azure AI Inference SDK

using Azure;
using Azure.AI.Inference;

ChatCompletionsClient client = new ChatCompletionsClient(
    new Uri("https://<resource>.services.ai.azure.com/models"),
    new AzureKeyCredential(Environment.GetEnvironmentVariable("AZURE_INFERENCE_CREDENTIAL"))
);

使用 Microsoft Entra ID 認證：

OpenAI 開發套件

using Azure.Identity;
using OpenAI;
using OpenAI.Chat;
using System.ClientModel.Primitives;

#pragma warning disable OPENAI001

BearerTokenPolicy tokenPolicy = new(
    new DefaultAzureCredential(),
    "https://ai.azure.com/.default"
);

ChatClient client = new(
    model: "gpt-4o-mini", // Your deployment name
    authenticationPolicy: tokenPolicy,
    options: new OpenAIClientOptions() {
        Endpoint = new Uri("https://<resource>.openai.azure.com/openai/v1/")
    }
);

Azure AI Inference SDK

using Azure;
using Azure.Identity;
using Azure.AI.Inference;

ChatCompletionsClient client = new ChatCompletionsClient(
    new Uri("https://<resource>.services.ai.azure.com/models"),
    new DefaultAzureCredential()
);

聊天完成

OpenAI 開發套件

using OpenAI.Chat;

ChatCompletion completion = client.CompleteChat(
    new SystemChatMessage("You are a helpful assistant."),
    new UserChatMessage("What is Azure AI?")
);

Console.WriteLine(completion.Content[0].Text);

Azure AI Inference SDK

using Azure.AI.Inference;

ChatCompletionsOptions requestOptions = new ChatCompletionsOptions()
{
    Messages = {
        new ChatRequestSystemMessage("You are a helpful assistant."),
        new ChatRequestUserMessage("What is Azure AI?")
    },
    Model = "DeepSeek-V3.1", // Optional for single-model endpoints
};

Response<ChatCompletions> response = client.Complete(requestOptions);
Console.WriteLine(response.Value.Choices[0].Message.Content);

串流

OpenAI 開發套件

using OpenAI.Chat;

CollectionResult<StreamingChatCompletionUpdate> updates = client.CompleteChatStreaming(
    new SystemChatMessage("You are a helpful assistant."),
    new UserChatMessage("Write a poem about Azure.")
);

foreach (StreamingChatCompletionUpdate update in updates)
{
    foreach (ChatMessageContentPart part in update.ContentUpdate)
    {
        Console.Write(part.Text);
    }
}

Azure AI Inference SDK

using Azure.AI.Inference;

ChatCompletionsOptions requestOptions = new ChatCompletionsOptions()
{
    Messages = {
        new ChatRequestSystemMessage("You are a helpful assistant."),
        new ChatRequestUserMessage("Write a poem about Azure.")
    },
    Model = "gpt-4o-mini",
};

StreamingResponse<StreamingChatCompletionsUpdate> response = client.CompleteStreaming(requestOptions);

await foreach (StreamingChatCompletionsUpdate update in response)
{
    if (update.ContentUpdate != null)
    {
        Console.Write(update.ContentUpdate);
    }
}

嵌入技術

OpenAI 開發套件

using OpenAI;
using OpenAI.Embeddings;
using System.ClientModel;

EmbeddingClient client = new(
    "text-embedding-3-small",
    credential: new ApiKeyCredential("API-KEY"),
    options: new OpenAIClientOptions()
    {

        Endpoint = new Uri("https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1")
    }
);

string input = "This is a test";

OpenAIEmbedding embedding = client.GenerateEmbedding(input);
ReadOnlyMemory<float> vector = embedding.ToFloats();
Console.WriteLine($"Embeddings: [{string.Join(", ", vector.ToArray())}]");

Azure AI Inference SDK

using Azure;
using Azure.AI.Inference;

EmbeddingsClient client = new EmbeddingsClient(
    new Uri("https://<resource>.services.ai.azure.com/models"),
    new AzureKeyCredential(Environment.GetEnvironmentVariable("AZURE_INFERENCE_CREDENTIAL"))
);

EmbeddingsOptions embeddingsOptions = new EmbeddingsOptions()
{
    Input = { "Your text string goes here" },
    Model = "text-embedding-3-small"
};

Response<EmbeddingsResult> response = client.Embed(embeddingsOptions);
ReadOnlyMemory<float> embedding = response.Value.Data[0].Embedding;

設定

安裝 OpenAI SDK：

npm install openai

若要進行 Microsoft Entra ID 認證，請同時安裝：

npm install @azure/identity

用戶端組態

使用 API 金鑰驗證：

OpenAI 開發套件

import { OpenAI } from "openai";

const client = new OpenAI({
    baseURL: "https://<resource>.openai.azure.com/openai/v1/",
    apiKey: process.env.AZURE_OPENAI_API_KEY
});

Azure AI Inference SDK

import ModelClient from "@azure-rest/ai-inference";
import { AzureKeyCredential } from "@azure/core-auth";

const client = ModelClient(
    "https://<resource>.services.ai.azure.com/models", 
    new AzureKeyCredential(process.env.AZURE_INFERENCE_CREDENTIAL)
);

使用 Microsoft Entra ID 認證：

OpenAI 開發套件

import { DefaultAzureCredential, getBearerTokenProvider } from "@azure/identity";
import { OpenAI } from "openai";

const tokenProvider = getBearerTokenProvider(
    new DefaultAzureCredential(),
    'https://ai.azure.com/.default'
);

const client = new OpenAI({
    baseURL: "https://<resource>.openai.azure.com/openai/v1/",
    apiKey: tokenProvider
});

Azure AI Inference SDK

import ModelClient from "@azure-rest/ai-inference";
import { DefaultAzureCredential } from "@azure/identity";

const clientOptions = { 
    credentials: { 
        scopes: ["https://cognitiveservices.azure.com/.default"] 
    } 
};

const client = ModelClient(
    "https://<resource>.services.ai.azure.com/models", 
    new DefaultAzureCredential(),
    clientOptions
);

聊天完成

OpenAI 開發套件

const completion = await client.chat.completions.create({
    model: "DeepSeek-V3.1", // Required: your deployment name
    messages: [
        { role: "system", content: "You are a helpful assistant." },
        { role: "user", content: "What is Azure AI?" }
    ]
});

console.log(completion.choices[0].message.content);

Azure AI Inference SDK

const response = await client.path("/chat/completions").post({
    body: {
        messages: [
            { role: "system", content: "You are a helpful assistant." },
            { role: "user", content: "What is Azure AI?" }
        ],
        model: "DeepSeek-V3.1" // Optional for single-model endpoints
    }
});

console.log(response.body.choices[0].message.content);

串流

OpenAI 開發套件

const stream = await client.chat.completions.create({
    model: "DeepSeek-V3.1",
    messages: [
        { role: "system", content: "You are a helpful assistant." },
        { role: "user", content: "Write a poem about Azure." }
    ],
    stream: true
});

for await (const chunk of stream) {
    if (chunk.choices[0]?.delta?.content) {
        process.stdout.write(chunk.choices[0].delta.content);
    }
}

Azure AI Inference SDK

const response = await client.path("/chat/completions").post({
    body: {
        messages: [
            { role: "system", content: "You are a helpful assistant." },
            { role: "user", content: "Write a poem about Azure." }
        ],
        model: "DeepSeek-V3.1",
        stream: true
    }
}).asNodeStream();

for await (const chunk of response) {
    if (chunk.choices && chunk.choices[0]?.delta?.content) {
        process.stdout.write(chunk.choices[0].delta.content);
    }
}

嵌入技術

OpenAI 開發套件

import OpenAI from "openai";
import { getBearerTokenProvider, DefaultAzureCredential } from "@azure/identity";

const tokenProvider = getBearerTokenProvider(
    new DefaultAzureCredential(),
    'https://ai.azure.com/.default');
const client = new OpenAI({
    baseURL: "https://<resource>.openai.azure.com/openai/v1/",
    apiKey: tokenProvider
});

const embedding = await client.embeddings.create({
  model: "text-embedding-3-large", // Required: your deployment name
  input: "The quick brown fox jumped over the lazy dog",
  encoding_format: "float",
});

console.log(embedding);

Azure AI Inference SDK

import ModelClient from "@azure-rest/ai-inference";
import { AzureKeyCredential } from "@azure/core-auth";

const client = ModelClient(
    "https://<resource>.services.ai.azure.com/models",
    new AzureKeyCredential(process.env.AZURE_INFERENCE_CREDENTIAL)
);

const response = await client.path("/embeddings").post({
    body: {
        input: ["Your text string goes here"],
        model: "text-embedding-3-small"
    }
});

const embedding = response.body.data[0].embedding;

設定

把 OpenAI SDK 加入你的 project。 請參考 OpenAI Java GitHub repository 以獲得最新版本與安裝說明。

關於 Microsoft Entra ID 認證，也新增：

<dependency>
    <groupId>com.azure</groupId>
    <artifactId>azure-identity</artifactId>
    <version>1.18.0</version>
</dependency>

用戶端組態

使用 API 金鑰驗證：

OpenAI 開發套件

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;

OpenAIClient client = OpenAIOkHttpClient.builder()
    .baseUrl("https://<resource>.openai.azure.com/openai/v1/")
    .apiKey(System.getenv("AZURE_OPENAI_API_KEY"))
    .build();

Azure AI Inference SDK

import com.azure.ai.inference.ChatCompletionsClient;
import com.azure.ai.inference.ChatCompletionsClientBuilder;
import com.azure.core.credential.AzureKeyCredential;

ChatCompletionsClient client = new ChatCompletionsClientBuilder()
    .credential(new AzureKeyCredential(System.getenv("AZURE_INFERENCE_CREDENTIAL")))
    .endpoint("https://<resource>.services.ai.azure.com/models")
    .buildClient();

使用 Microsoft Entra ID 認證：

OpenAI 開發套件

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.azure.identity.DefaultAzureCredential;
import com.azure.identity.DefaultAzureCredentialBuilder;

DefaultAzureCredential tokenCredential = new DefaultAzureCredentialBuilder().build();

OpenAIClient client = OpenAIOkHttpClient.builder()
    .baseUrl("https://<resource>.openai.azure.com/openai/v1/")
    .credential(BearerTokenCredential.create(
        AuthenticationUtil.getBearerTokenSupplier(
            tokenCredential, 
            "https://ai.azure.com/.default"
        )
    ))
    .build();

Azure AI Inference SDK

import com.azure.ai.inference.ChatCompletionsClient;
import com.azure.ai.inference.ChatCompletionsClientBuilder;
import com.azure.identity.DefaultAzureCredential;
import com.azure.identity.DefaultAzureCredentialBuilder;
import com.azure.core.credential.TokenCredential;

TokenCredential credential = new DefaultAzureCredentialBuilder().build();
ChatCompletionsClient client = new ChatCompletionsClientBuilder()
    .credential(credential)
    .endpoint("https://<resource>.services.ai.azure.com/models")
    .buildClient();

聊天完成

OpenAI 開發套件

import com.openai.models.chat.completions.*;

ChatCompletionCreateParams params = ChatCompletionCreateParams.builder()
    .addSystemMessage("You are a helpful assistant.")
    .addUserMessage("What is Azure AI?")
    .model("DeepSeek-V3.1") // Required: your deployment name
    .build();

ChatCompletion completion = client.chat().completions().create(params);
System.out.println(completion.choices().get(0).message().content());

Azure AI Inference SDK

import com.azure.ai.inference.models.*;
import java.util.List;

List<ChatRequestMessage> messages = List.of(
    new ChatRequestSystemMessage("You are a helpful assistant."),
    new ChatRequestUserMessage("What is Azure AI?")
);

ChatCompletionsOptions options = new ChatCompletionsOptions(messages);
options.setModel("DeepSeek-V3.1"); // Optional for single-model endpoints

ChatCompletions response = client.complete(options);
System.out.println(response.getChoices().get(0).getMessage().getContent());

串流

OpenAI 開發套件

import com.openai.models.chat.completions.*;
import java.util.stream.Stream;

ChatCompletionCreateParams params = ChatCompletionCreateParams.builder()
    .addSystemMessage("You are a helpful assistant.")
    .addUserMessage("Write a poem about Azure.")
    .model("DeepSeek-V3.1") // Required: your deployment name
    .build();

Stream<ChatCompletionChunk> stream = client.chat().completions().createStreaming(params);

stream.forEach(chunk -> {
    if (chunk.choices() != null && !chunk.choices().isEmpty()) {
        String content = chunk.choices().get(0).delta().content();
        if (content != null) {
            System.out.print(content);
        }
    }
});

Azure AI Inference SDK

import com.azure.ai.inference.models.*;

List<ChatRequestMessage> messages = List.of(
    new ChatRequestSystemMessage("You are a helpful assistant."),
    new ChatRequestUserMessage("Write a poem about Azure.")
);

ChatCompletionsOptions options = new ChatCompletionsOptions(messages);
options.setModel("DeepSeek-V3.1");

IterableStream<ChatCompletions> response = client.completeStream(options);

response.forEach(update -> {
    if (update.getChoices() != null && !update.getChoices().isEmpty()) {
        String content = update.getChoices().get(0).getDelta().getContent();
        if (content != null) {
            System.out.print(content);
        }
    }
});

嵌入技術

OpenAI 開發套件

package com.openai.example;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.embeddings.EmbeddingCreateParams;
import com.openai.models.embeddings.EmbeddingModel;

public final class EmbeddingsExample {
    private EmbeddingsExample() {}

    public static void main(String[] args) {
        // Configures using one of:
        // - The `OPENAI_API_KEY` environment variable
        // - The `OPENAI_BASE_URL` and `AZURE_OPENAI_KEY` environment variables
        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        EmbeddingCreateParams createParams = EmbeddingCreateParams.builder()
                .input("The quick brown fox jumped over the lazy dog")
                .model(EmbeddingModel.TEXT_EMBEDDING_3_SMALL)
                .build();

        System.out.println(client.embeddings().create(createParams));
    }
}

Azure AI Inference SDK

import com.azure.ai.inference.EmbeddingsClient;
import com.azure.ai.inference.EmbeddingsClientBuilder;
import com.azure.core.credential.AzureKeyCredential;

EmbeddingsClient client = new EmbeddingsClientBuilder()
    .credential(new AzureKeyCredential(System.getenv("AZURE_INFERENCE_CREDENTIAL")))
    .endpoint("https://<resource>.services.ai.azure.com/models")
    .buildClient();

EmbeddingsOptions embeddingsOptions = new EmbeddingsOptions(
    List.of("Your text string goes here")
);
embeddingsOptions.setModel("text-embedding-3-small");

EmbeddingsResult response = client.embed(embeddingsOptions);
List<Float> embedding = response.getData().get(0).getEmbedding();

設定

安裝 OpenAI SDK：

go get github.com/openai/openai-go/v3

若要進行 Microsoft Entra ID 認證，請同時安裝：

go get -u github.com/Azure/azure-sdk-for-go/sdk/azidentity

用戶端組態

使用 API 金鑰驗證：

OpenAI 開發套件

import (
    "github.com/openai/openai-go/v3"
    "github.com/openai/openai-go/v3/option"
)

client := openai.NewClient(
    option.WithBaseURL("https://<resource>.openai.azure.com/openai/v1/"),
    option.WithAPIKey(os.Getenv("AZURE_OPENAI_API_KEY")),
)

Azure AI Inference SDK

Azure AI Inference SDK for Go 使用 Azure SDK 模式。

使用 Microsoft Entra ID 認證：

OpenAI 開發套件

import (
    "github.com/Azure/azure-sdk-for-go/sdk/azidentity"
    "github.com/openai/openai-go/v3"
    "github.com/openai/openai-go/v3/azure"
    "github.com/openai/openai-go/v3/option"
)

tokenCredential, err := azidentity.NewDefaultAzureCredential(nil)
if err != nil {
    panic(err)
}

client := openai.NewClient(
    option.WithBaseURL("https://<resource>.openai.azure.com/openai/v1/"),
    azure.WithTokenCredential(tokenCredential),
)

Azure AI Inference SDK

Azure AI Inference SDK for Go 支援 Microsoft Entra ID 透過 Azure SDK。

聊天完成

OpenAI 開發套件

import (
    "context"
    "fmt"
    "github.com/openai/openai-go/v3"
)

chatCompletion, err := client.Chat.Completions.New(context.TODO(), openai.ChatCompletionNewParams{
    Messages: []openai.ChatCompletionMessageParamUnion{
        openai.SystemMessage("You are a helpful assistant."),
        openai.UserMessage("What is Azure AI?"),
    },
    Model: "DeepSeek-V3.1", // Required: your deployment name
})

if err != nil {
    panic(err.Error())
}

fmt.Println(chatCompletion.Choices[0].Message.Content)

Azure AI Inference SDK

Azure AI Inference SDK for Go 使用 Azure SDK 模式來完成聊天應答。

串流

OpenAI 開發套件

import (
    "context"
    "fmt"
    "github.com/openai/openai-go/v3"
)

stream := client.Chat.Completions.NewStreaming(context.TODO(), openai.ChatCompletionNewParams{
    Messages: []openai.ChatCompletionMessageParamUnion{
        openai.SystemMessage("You are a helpful assistant."),
        openai.UserMessage("Write a poem about Azure."),
    },
    Model: "DeepSeek-V3.1", // Required: your deployment name
})

for stream.Next() {
    chunk := stream.Current()
    if len(chunk.Choices) > 0 && chunk.Choices[0].Delta.Content != "" {
        fmt.Print(chunk.Choices[0].Delta.Content)
    }
}

if err := stream.Err(); err != nil {
    panic(err.Error())
}

Azure AI Inference SDK

Azure AI Inference SDK for Go 支援透過 Azure SDK 模式進行串流。

嵌入技術

OpenAI 開發套件

package main

import (
    "context"
    "fmt"
    "log"

    "github.com/Azure/azure-sdk-for-go/sdk/azidentity"
    "github.com/openai/openai-go/v3"
    "github.com/openai/openai-go/v3/azure"
    "github.com/openai/openai-go/v3/option"
)

func main() {
    tokenCredential, err := azidentity.NewDefaultAzureCredential(nil)
    if err != nil {
        log.Fatalf("Error creating credential:%s", err)
    }
    // Create a client with Azure OpenAI endpoint and Entra ID credentials
    client := openai.NewClient(
        option.WithBaseURL("https://YOUR-RESOURCE-NAME.openai.azure.com/openai/v1/"),
        azure.WithTokenCredential(tokenCredential),
    )

    inputText := "The quick brown fox jumped over the lazy dog"

    // Make the embedding request synchronously
    resp, err := client.Embeddings.New(context.Background(), openai.EmbeddingNewParams{
        Model: openai.EmbeddingModel("text-embedding-3-large"), // Use your deployed model name on Azure
        Input: openai.EmbeddingNewParamsInputUnion{
            OfArrayOfStrings: []string{inputText},
        },
    })
    if err != nil {
        log.Fatalf("Failed to get embedding: %s", err)
    }

    if len(resp.Data) == 0 {
        log.Fatalf("No embedding data returned.")
    }

    // Print embedding information
    embedding := resp.Data[0].Embedding
    fmt.Printf("Embedding Length: %d\n", len(embedding))
    fmt.Println("Embedding Values:")
    for _, value := range embedding {
        fmt.Printf("%f, ", value)
    }
    fmt.Println()
}

Azure AI Inference SDK

Azure AI Inference SDK for Go 使用 Azure SDK pattern 來進行嵌入計算。

常見的移轉模式

模型參數處理

• Azure AI Inference SDK：model 參數對單模型端點為可選，但多模型端點必須。
• OpenAI SDK：參數 model 一律是必需的，而且應該設定為您的部署名稱。

端點 URL 格式

• Azure AI Inference SDK：使用 https://<resource>.services.ai.azure.com/models。
• OpenAI SDK：使用 https://<resource>.openai.azure.com/openai/v1（連接 OpenAI v1 API）。

回應結構

回應結構類似，但有一些差異：

• Azure AI 推理 SDK：回傳具有ChatCompletions的choices[].message.content物件。
• OpenAI SDK：傳回包含ChatCompletion的choices[].message.content物件。

這兩個 SDK 都提供類似的回應資料access模式，包括：

• 訊息內容
• 代幣使用
• 型號資訊
• 完成原因

移轉檢查清單

使用此檢查清單可確保順利移轉：

• 安裝適用於您程式設計語言的 OpenAI SDK
• 更新認證碼（API 金鑰或 Microsoft Entra ID）
• 將端點網址從 .services.ai.azure.com/models 改為 .openai.azure.com/openai/v1/
• 將資格範圍 https://cognitiveservices.azure.com/.default 從 改為 https://ai.azure.com/.default
• 更新用戶端初始化程式碼
• 請務必使用 model 您的部署名稱作為參數指定
• 更新要求方法呼叫 (complete → chat.completions.create)
• 更新串流程式碼 (如果適用的話)
• 更新錯誤處理以使用 OpenAI SDK 例外狀況
• 徹底測試所有功能
• 更新文件和程式碼註解

故障排除

驗證失敗

如果您遇到驗證失敗：

• 確認您的 API 金鑰正確且未過期
• 對於 Microsoft Entra ID，請確保您的應用程式擁有正確的權限
• 檢查證照範圍是否設為https://ai.azure.com/.default

端點錯誤

如果您收到端點錯誤：

• 確認端點 URL 格式包含 /openai/v1/ 作為結尾。
• 請確定您的資源名稱正確無誤。
• 檢查模型部署是否存在且處於作用中狀態。

模型未找到錯誤

如果您收到「找不到模型」的錯誤訊息：

• 確認您使用的是部署名稱，而不是模型名稱。
• 檢查部署是否在你的 Microsoft Foundry 資源中啟用。
• 請確定部署名稱完全相符 （區分大小寫）。

相關內容

• Azure OpenAI 支援程式語言
• 如何使用 Foundry 模型產生聊天完成
• API 演進和版本生命週期
• 在 OpenAI 和Azure OpenAI 端點之間切換