Today, we are available in Azure AI Foundry for the new flagship Openi, GPT-5. This is more than the new edition of the model; It is the strongest model of a large language (LLM) ever released across key benchmarks.
For the leading companies building with AI, the conversation shifted behind the cat. The bar is high: can your generation, reason and measurable results – safe and scale?
Today we announce general avilabitis The new flagship model OPENAI, GPT-5, in Azure Ai Foundry. This is more than the new edition of the model; It is the strongest LLM ever released across key benchmarks. GPT-5 in Azure AI Foundry Peirs Frontier Consistent with high-performance generation and cost-effective, delivered on Microsoft Azure’s business class platform, so organizations can move confidently from pilots for production.
GPT-5 in Azure AI Foundry: built for workload in the real world
In the Azure AI foundry, the GPT-5 models are available via the API and organized by the Temodel router. The GPT-5 series is expanding the additional Stregths:
- GPT-5The complete model of reasoning provides a deep and richer justification of analysts and complex tasks such as code generation, with a token context of 272K.
- GPT-5 mini Powering real -time experience for applications and agents that require justification, calling tools to solve customer problems.
- GPT-5 nano It is a new class of reasoning model that focuses on ultra-all and speed with the rich abilities of Q&A.
- Chat gpt-5 Natural, multimodal, multimodal, multiple rotation that remains in the context of consciousness in agent workflows, with a context of 128k token.
Together, the apartment brings a trouble -free continuum from strict agent coding tasks to relatively simple Q&A – all delivered with the same AZure AI foundry by using the model in the foundry models.
Under the hood of the GPT-5, it unifies advanced reasoning, code generation and the interaction of natural language. It combines an analytical depth with an intuitive dialogue to solve end-to-end problems and explains its approach. The agency enables multiple use of a tool with multiple and long action chains with transparent and audited decisions. As a model coding model, the GPT-5 can plan a comprehensive agent workflow, building migration and refactor code, as well as production and documentation with clear rational. Developer controls – including parameters reasoning_effort and eloquence -t teams tune in depth, speed and details while New Tool Call Functions FreeForm Enable the extension of the compatibility of the tool with rigid schemes.
Orchestrate with model router -Dset scale with agents
Introduction GPT-5 to Azure AI Foundry is more than a model drop: it’s a forward jump for a platform. From today, developers can use model router in founders to maximize the capabilities of the GPT-5 family models (and other models in foundry models) and save up to 60% for inference costs without losing loyalty. The model router, powered by a finely tuned SLM under the hood, evaluates every challenge and decides on the optimal model based on the complexity, performance needs and cost efficiency of each task. Allow the model to select the right model to easily build your Ai-Power applications.
And orchestration does not stop in routing – the foundry carries the same intelligence for agents. Soon, the GPT-5 will be available at Foundry Service Agent Agent, which pairs the Frontier models with built-in tools, including the integration of New browser automation and model context (MCP). Result: Policy controlled, use of tools that can search, act in web applications and complete the end-to-end tasks-instrumented with Foundry telemetry and matched to Microsoft AI.
Acceleration of business impact with GPT-5
These skills are mapped directly on the business impact.
In research and knowledge work, GPT-5 speeds up financial and legal analysis, market intelligence and reading properly diligence on a scale and produce the production prepared for decision-making. In operation and decision -making strengthens the logistics support, risks and processing of requirements by pairing robust reasoning with adherence to policy. Copilots and teams for customers experience benefit from multimodal agents who are reasonable in real time, call tools, solve tasks and return to people with a more useful context.
In GPT-5 software engineering, it excels in the generation of code, modernization of applications and high-quality engineering-glitting code style and explanation for compressing the inspection cycles.
And for the use of Whisses, which are sensitive to cost or latency, the architecture of Ultra Show-Lalence GPT-5-Nano provides fast high-precision centers, making it an ideal goal for a fine and high-volume model, direct requirements.
GPT-5 Customer Reflector
Customers release GPT-5 across complex, critical workload, which increases decision-making, supercharging coding and catalyzing product innovation.
Sap
SAP is excited to be among the first to use the GPT-5 power in Azure AI Foundry in our generative AI mushroom in the AA Foundation. The GPT-5 in the Azure AI Foundry will allow our product team to provide impressive business innovations to our customers.
—Dr. Walter Sun, please and global leading AI, SAP is
Relativity
The GPT-5 in the Azure AI Foundry increases the Latka to bring the intelligence of legal data into the action … This AI next generation seizes legal teams to reveal deeper knowledge, speed up decision making and increase stronger strategies throughout the legal process.
—Dr. ARON AHMADIA, Head of Director, Applied Science, Relativity
Weekly
The partnership between Hebbia and Azure AI Foundry gives financial professionals an unprecedented advantage. With the advanced justification of the GPT-5 in Hebbia, critical characters can determine across thousands of documents and structure a comprehensive financial analysis with speed and accuracy.
—Danny Wheler, vice president of business and strategy
Building with AI in Github Copilot and Visual Studio
GPT-5 will start expanding to Milles of Developers today using Github Copilot and Visual Studio CodeApplication of advanced capabilities of justifying the flagship model to increase complex problems -a sophisticated refractoring for more efficient navigation in large code directions. GPT-5 helps developers to write, test and deploy code faster while supporting agent coding tasks with significant improvisations for the coding style and overall quality code. With GPT-5 developers not only code faster but also code.
With today’s VS code, developers also gain a stronger experience with the agent coding directly in the editor: the coding agent Github Copilot has an improved experience for autonomous tasks in the background. In addition, Github Copilot Chat brings increased productivity, including support beyond 128 tools for a single chat requirement and cottage checkpoints, allowing users to restore changes in the workspace in the previous point. Today, we also announce updated extensions for the development of agents using Azure AI Foundry All in VS Codes.
These announcements are expanding Microsoft’s strategy to transform the development of software S AI and bring advanced AI capacities to the entire software life cycle.
Safety, Safety and Administration as designed
In all Domans, the safety and safety of a layering cake of protection, which together provide the protection of risk scenarios – and AI does not differ. For AI we think about layers with a model as a core. With the GPT-5, the core is safer than before:
The Microsoft AI Red team found that the GPT-5 has one of the most modern security profiles of any Openai model that performs S-or better than-O3.
—Dr. Sarah Bird, Chief Product Director responsible AI, Microsoft
As we think about layers of safety, security and management, they surround this core –Azure AI Foundry Provides a number of other controls:
- Azure AI content safety The protection is applied to each challenge and completion, such as fast shields that help detect and alleviate attempts to injection fast injection before reaching the model.
- Built -in evaluators work with AI AII Agent AI Red Make alignment, bias and safety tests during development and production, while continuous evaluation Metric metrics in real time, quality, safety and justice stream to the azure monitor and application knowledge for visibility with one panel.
- Safety signals are finally integrated directly with Microsoft Defender for Cloudand Runtime metadata and evaluation results are integrated Microsoft PurView For audit, prevent data loss and regulatory reporting, prolongation of protection and management throughout the GPT-5 life cycle.
Bringing AI to every workflow with Github Copilot and Visual Studio Code
From today, GPT-5 begins to introduce Millions of developers Use Github Copilot and the Visual Studio code that will be able to Select GPT-5 to write, test and deploy code—A and developing agents using Azure AI Foundry All in the VS code. Comprehensive agent coding GPT-5 with meaningful improvisations for personality coding, front-end aesthetics and quality code, highly required improvisation for the development community.
Our evaluation shows OPENAI GPT-5 capabilities and O3 contextual consciousness excess, which allows developers to solve more complicated factoring problems for navigating large code foundations. For the GPT-5, users in the Visual Studio family can not only encode faster but better encode.
Vs code and our A recent decision on Open-Source Github CopilotIt represents our commitment to open tools and standards and demonstrates our ability to meet the rapid pace of innovation of the model while maintaining developers at the forefront. In today’s issue of VS Code, developers can. IN Today’s VS version codeDevelopers have even more control over their experience with chat – enhancement of weld with terminal tool binding, collector updates and tool limits, new checkpoints and more.
Today’s Ondon expands Microsoft’s strategy to transform the development of software with AI and brings advanced AI capacity to the entire software life cycle.
Start building today
The GPT-5 is available through our standard Azure and Azure offer, while the selection is optimized for cost efficiency and control needs, including global and data zones (United States, European Union).1
Thanks to the connection of the Azure AI Foundry class, real-time rating, built-in vitality and secure deployment options, you can surely move from Prot to production-VŠ, while unique tools such as the router optimizes quality, latency and costs across workload.
1Prices is an accident from August 2025
(Tagstotranslate) ai