Implementation:Microsoft Semantic kernel AddOpenAIChatClient

Knowledge Sources	Semantic Kernel Semantic Kernel Docs
Domains	AI_Orchestration, Service_Connector_Architecture
Last Updated	2026-02-11 19:00 GMT

Overview

Concrete tool for registering an OpenAI chat completion service with the kernel builder provided by the Microsoft Semantic Kernel library.

Description

AddOpenAIChatClient is an extension method on IKernelBuilder that registers an OpenAI-backed chat completion client into the kernel's dependency injection container. It configures the connection to OpenAI's API using the specified model identifier and API key, and optionally accepts an organization ID, a service ID for multi-service scenarios, and a custom HttpClient for network-level customization.

Under the hood, this method creates an OpenAIClient with the provided credentials and registers it as an IChatClient (from the Microsoft.Extensions.AI abstraction layer) in the builder's IServiceCollection. This registration follows the Microsoft.Extensions.AI client pattern, which provides a standardized interface over multiple AI providers.

Usage

Use this method when your application needs to interact with OpenAI models for chat completion. Call it on the IKernelBuilder returned by Kernel.CreateBuilder() before calling Build(). If your application uses Azure-hosted OpenAI models, use the Azure-specific variant instead.

Code Reference

Source Location

Repository: semantic-kernel
File: dotnet/src/Connectors/Connectors.OpenAI/Extensions/OpenAIKernelBuilderExtensions.ChatClient.cs:L28-70

Signature

public static IKernelBuilder AddOpenAIChatClient(
    this IKernelBuilder builder,
    string modelId,
    string apiKey,
    string? orgId = null,
    string? serviceId = null,
    HttpClient? httpClient = null)

Import

using Microsoft.SemanticKernel;

I/O Contract

Inputs

Name	Type	Required	Description
builder	`IKernelBuilder`	Yes	The kernel builder instance to register the service on (implicit via extension method).
modelId	`string`	Yes	The OpenAI model identifier (e.g., "gpt-4o", "gpt-4o-mini").
apiKey	`string`	Yes	The API key for authenticating with OpenAI.
orgId	`string?`	No	Optional OpenAI organization identifier for billing and access control.
serviceId	`string?`	No	Optional unique identifier for this service registration, used to distinguish between multiple AI services in the same kernel.
httpClient	`HttpClient?`	No	Optional custom HttpClient for controlling timeouts, proxies, and other HTTP-level behavior.

Outputs

Name	Type	Description
return	`IKernelBuilder`	The same builder instance, enabling fluent method chaining.

Usage Examples

Basic Registration

using Microsoft.SemanticKernel;

Kernel kernel = Kernel.CreateBuilder()
    .AddOpenAIChatClient(
        modelId: "gpt-4o",
        apiKey: "your-api-key")
    .Build();

Registration with Organization ID

using Microsoft.SemanticKernel;

Kernel kernel = Kernel.CreateBuilder()
    .AddOpenAIChatClient(
        modelId: "gpt-4o",
        apiKey: "your-api-key",
        orgId: "org-abc123")
    .Build();

Multi-Service Registration with Service IDs

using Microsoft.SemanticKernel;

Kernel kernel = Kernel.CreateBuilder()
    .AddOpenAIChatClient(
        modelId: "gpt-4o",
        apiKey: "your-api-key",
        serviceId: "primary")
    .AddOpenAIChatClient(
        modelId: "gpt-4o-mini",
        apiKey: "your-api-key",
        serviceId: "fast")
    .Build();

Registration with Custom HttpClient

using Microsoft.SemanticKernel;
using System.Net.Http;

var httpClient = new HttpClient { Timeout = TimeSpan.FromSeconds(120) };

Kernel kernel = Kernel.CreateBuilder()
    .AddOpenAIChatClient(
        modelId: "gpt-4o",
        apiKey: "your-api-key",
        httpClient: httpClient)
    .Build();

Related Pages

Implements Principle

Principle:Microsoft_Semantic_kernel_AI_Service_Registration

Requires Environment

Uses Heuristic

Heuristic:Microsoft_Semantic_kernel_Custom_HttpClient_Retry_Policy

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment