Microsoft has unveiled a new era in AI development with the launch of Azure AI Foundry, a full-stack AI app and agent factory that empowers developers to create with AI. The platform provides a unified space for building, fine-tuning, and deploying intelligent agents, both in the cloud and on edge devices.
At the heart of this vision is OpenAI’s gpt-oss (open-weight) models, which are now available on Azure AI Foundry. These models offer unprecedented flexibility and control, allowing developers to run, adapt, and deploy OpenAI models entirely on their own terms.
The two flagship models, gpt-oss-120b and gpt-oss-20b, cater to different use cases. The former is a reasoning powerhouse, excelling at complex tasks like math, code, and domain-specific Q&A. In contrast, the latter is tool-savvy and lightweight, optimized for agentic tasks like code execution and tool use.
With Azure AI Foundry, developers can:
Spin up inference endpoints using gpt-oss in the cloud with just a few CLI commands
Fine-tune and distill the models using their own data and deploy with confidence
Mix open and proprietary models to match task-specific needs
The platform also introduces Foundry Local, which brings prominent open-source models to Windows AI Foundry, pre-optimized for inference on your own hardware. This enables organizations to deploy gpt-oss-20b on modern high-performance Windows PCs, without compromising security or data ownership.
By making these cutting-edge models available through an open platform, Microsoft is democratizing access to AI and empowering developers, decision-makers, and businesses to innovate with confidence. The company’s vision for the future is one of ubiquitous AI, where innovation is driven by openness and responsibility.
Source: https://azure.microsoft.com/en-us/blog/openais-open%E2%80%91source-model-gpt%E2%80%91oss-on-azure-ai-foundry-and-windows-ai-foundry