Microsoft and OpenAI recently announced the general availability (GA) of OpenAI’s GPT-5 model suite within the Azure AI Foundry platform, a unified platform for enterprise AI operations. The new model suite is designed for a range of specific workloads and optimized to handle real-world applications.
According to Microsoft CEO Satya Nadella’s comment on LinkedIn:
Today, GPT-5 launches across our platforms, including Microsoft 365 Copilot, Copilot, GitHub Copilot, and Azure AI Foundry. It’s the most capable model yet from our partners at OpenAI, bringing powerful new advances in reasoning, coding, and chat, all trained on Azure.
A core innovation of GPT-5 is its design as an internal orchestrator that automatically routes tasks to a team of specialized sub-models or “experts”, which is an approach that allows the model to handle different parts of a single prompt, such as text comprehension, planning, and creative writing, by switching between these experts behind the scenes.
Dharita Chokshi, a senior product manager at Walmart Global Tech India, wrote in a Medium article that this design leads to higher quality outputs, fewer hallucinations, and less need for extensive prompt tuning.
The GPT-5 models are available via API and orchestrated by the model router (preview) in Azure AI Foundry. The models includes:
- GPT-5, which is a full reasoning model with a large 272k token context, intended for complex tasks like analytics and code generation.
- GPT-5 mini that powers real-time applications and agents that require reasoning and tool-calling capabilities.
- GPT-5 nano, which is a new class of reasoning model focused on ultra-low latency and speed for high-volume, straightforward requests and fine-tuning.
- GPT-5 chat, which enables natural, multimodal, multi-turn conversations with a 128k token context.
The company emphasizes that the models are built for real-world workloads. As Steve Sweetman, an Azure OpenAI service product lead, explains:
Under the hood, GPT-5 unifies advanced reasoning, code generation, and natural language interaction. It combines analytical depth with intuitive dialogue to solve end-to-end problems and explain its approach.
Currently, the new models are being integrated into developer tools like GitHub Copilot and Visual Studio Code, offering enhanced reasoning for complex problems such as refactoring and navigating large codebases.
(Source: Microsoft AI and Machine Learning blog post)
In another Medium article, the authors conclude:
With GPT-5 now available in Azure AI Foundry, Microsoft is positioning the platform as a core enabler of enterprise AI transformation, offering the tools to move beyond experimentation and into scalable, impactful AI deployment.
Lastly, GPT-5 is available via Microsoft’s Standard offering in Azure AI Foundry, with deployment choices optimized for cost-efficiency and governance needs, including Global and Data Zone (United States, European Union) deployment options for data residency and compliance.
