Steve Sweetman announces the general availability of GPT-5 in Azure AI Foundry, detailing its rollout to developers via GitHub Copilot and VS Code, enhanced reasoning, orchestration features, and robust security and governance layers for enterprise AI adoption.

GPT-5 in Azure AI Foundry: The Future of AI Apps and Agents Starts Here

By Steve Sweetman

Microsoft introduces GPT-5, OpenAI’s most advanced large language model, now generally available in Azure AI Foundry. This model sets new standards across key AI benchmarks, providing richer reasoning, higher performance, and enterprise-grade reliability.

Key Highlights

  • Multiple GPT-5 Variants: Includes GPT-5, GPT-5 mini, GPT-5 nano, and GPT-5 chat, each optimized for tasks ranging from deep analytics to real-time Q&A and agentic workflows. Context window support up to 272k tokens.
  • Business and Coding Revolution: GPT-5 is designed not just for conversations, but for generating reasoning-led outcomes, advanced analytics, and complex coding tasks—directly benefiting teams using GitHub Copilot and Visual Studio Code.
  • Model Router Orchestration: Developers orchestrate workloads using the model router, which intelligently allocates tasks across models based on complexity, speed, and cost.
  • End-to-End Agents: Foundry Agents combine GPT-5 with browser automation, tool calling, and Model Context Protocol (MCP), supporting multi-step workflows, transparency, and compliance.

Technical and Business Impact

  • Software Engineering: Accelerated code generation, modernization, review cycles, and documentation, with significant improvements to coding quality and agentic capabilities. Advanced developer controls like reasoning_effort and tool-calling provide fine-tuned performance.
  • Real-World Workloads: Case studies from SAP, Relativity, and Hebbia illustrate GPT-5’s use for legal, financial, and business analysis, powering mission-critical workloads and delivering actionable insights at scale.
  • Agentic Coding in VS Code: Developers benefit from improved agentic experiences inside Visual Studio Code, with GitHub Copilot supporting more complex, autonomous coding and chat workflows. Over 128 tools now supported per chat, with checkpointing and workspace restoration.

Security, Safety, and Compliance by Design

  • Layered Protections: GPT-5 models include vetted safeguards such as Azure AI Content Safety, prompt shields, agent evaluators, and continuous runtime evaluation.
  • Integration with Security Platforms: Evaluation results connect with Microsoft Defender for Cloud and Microsoft Purview, providing real-time oversight and streamlining compliance and auditing.
  • Responsible AI Commitment: Built-in alignment testing, risk monitoring, and reporting ensure deployments meet Microsoft’s Responsible AI standards.

Deployment and Getting Started

  • Flexible Offerings: Multiple deployment zones and options for cost efficiency, global scale, and compliance requirements.
  • Reliability and Monitoring: Azure AI Foundry delivers real-time observability, built-in monitoring, and secure production readiness, helping organizations swiftly move from pilot to production.

Customer Endorsements:

  • SAP: Empowering developer communities in their generative AI hub.
  • Relativity: Enhancing legal data intelligence and decision-making with GPT-5.
  • Hebbia: Enabling fast and accurate financial analysis at scale.

Images and Diagrams

  • Screenshots and architectural diagrams of GPT-5 agentic workflows, VS Code integration, and Foundry model routing.

For organizations and developers looking to advance with AI, GPT-5 in Azure AI Foundry offers a unified, secure, and highly capable platform to drive the next generation of intelligent applications.

This post appeared first on “The Azure Blog”. Read the entire article here