Jon Galloway, Pamela Fox, and Dan Wahlin deliver a deep dive into the integration of GPT-5 across GitHub Copilot, Azure AI, and core Microsoft developer tools, offering practical code samples and guidance for immediate adoption.

GPT-5 Integrations for Microsoft Developers: GitHub Copilot, Azure AI, and VS Code

Authors: Jon Galloway, Pamela Fox, Dan Wahlin

OpenAI’s GPT-5 model has landed, and Microsoft’s developer ecosystem now features GPT-5 integrations across GitHub Copilot, Visual Studio Code, Azure AI Foundry, Copilot Studio, and more. This article provides a dev-focused roundup: how to access GPT-5, what’s new, and how to start building with it immediately.

GPT-5 Overview

Improved reasoning and accuracy for complex workflows
Unified capabilities (chat, agents, coding, multimodal, advanced math)
Faster response times and larger context handling
Available in API and developer tools

Where You Can Use GPT-5 Today

GitHub Copilot

Leverages GPT-5 for richer, longer code completions and intelligent suggestions
Integrates with VS Code, Visual Studio, JetBrains, Xcode, and Eclipse (preview availability varies)
Accessible in all Copilot plans, with GPT-5 mini as a starting point

Visual Studio Code (AI Toolkit)

Experiment with GPT-5 via the AI Toolkit
Connects to GitHub Models, Azure AI Foundry, and supports both cloud and open-source/local backends
Official announcement

Azure AI Foundry

Enterprise-grade, secure access to GPT-5 (requires registration for main model; mini/nano/chat variants are available without registration)
Regional availability: East US 2, Sweden Central
Supports long-running agents and structured outputs (details)

GitHub Models Marketplace

Supports multiple GPT-5 variants for rapid experiment and workflow integration
Marketplace link

Microsoft Copilot Studio

Makers can use GPT-5 models for orchestrating agents and building advanced chat/automation flows

OpenAI .NET SDK, Python, JavaScript

Official SDKs for .NET, Python, and JavaScript enable GPT-5 integration in C#, Python, and JS/TypeScript apps with new API features

Sample: C# (Streaming, Reasoning)

OpenAIResponseClient client = new(model: "gpt-5", apiKey: Environment.GetEnvironmentVariable("OPENAI_API_KEY"));
await foreach (var update in client.CreateResponseStreamingAsync(
    userInputText: "Explain beta-reduction in lambda calculus.",
    new ResponseCreationOptions { ReasoningOptions = new ResponseReasoningOptions { ReasoningEffortLevel = ResponseReasoningEffortLevel.High } }))
{
    if (update is StreamingResponseContentPartDeltaUpdate delta) {
        Console.Write(delta.Text);
    }
}

Sample: Python (Controllable Reasoning & Verbosity)

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

client = openai.AzureOpenAI(
    api_version=os.environ["AZURE_OPENAI_VERSION"],
    azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
    azure_ad_token_provider=get_bearer_token_provider(DefaultAzureCredential(), "https://cognitiveservices.azure.com/.default")
)
response = client.chat.completions.create(
    model=os.environ["AZURE_OPENAI_DEPLOYMENT"],
    messages=[{"role": "user", "content": "Explain beta-reduction in lambda calculus."}],
    reasoning_effort="minimal",
    verbosity="low"
)
print(response.choices[0].message.content)

Sample: JavaScript (Structured Output, Azure Foundry Integration)

import { AzureOpenAI } from "openai";
import dotenv from "dotenv";
dotenv.config();
const client = new AzureOpenAI({
    endpoint: process.env.AZURE_INFERENCE_ENDPOINT,
    apiKey: process.env.AZURE_INFERENCE_KEY,
    apiVersion: "2025-01-01-preview",
    deployment: process.env.AZURE_OPENAI_DEPLOYMENT || "gpt-5",
});
const schema = { /* custom JSON schema as in full article */ };
const result = await client.chat.completions.create({
    model: deployment,
    messages: [{ role: "system", content: "Return JSON only." }, { role: "user", content: "What is 23 * 7? Show your steps." }],
    response_format: { type: "json_schema", json_schema: schema },
});
const data = JSON.parse(result.choices[0].message?.content ?? "{}" );
console.log("Steps:", data.steps);
console.log("Answer:", data.answer);

DevOps & Evaluation

Use the Azure AI Evaluation SDK for model quality and cost assessment
Try model variants in side-by-side playgrounds via Azure AI Foundry or GitHub Models
.NET apps: Ensure evaluation consistency with Microsoft.Extensions.AI Evaluation

Real-World Examples & Community

Pamela Fox: Deep dive into GPT-5 for RAG (retrieval-augmented generation)
Anthony Shaw: Updates GitHub Models CLI for GPT-5 family and explores LLM use in GitHub Actions for automated PR summaries, code review, and CI workflows
Burke Holland: “Vibe coding” experiments in VS Code—GPT-5 builds a fully working website and game

Resources & Quick Links

Microsoft’s dev ecosystem is GPT-5 ready, enabling fast, scalable adoption for coding, DevOps, AI applications, and more.

This post appeared first on “Microsoft DevBlog”. Read the entire article here