Microsoft used NVIDIA GTC on March 16, 2026 to widen Microsoft Foundry and Azure AI in three directions: production agent tooling, next-generation NVIDIA infrastructure, and Physical AI workflows. The company said Foundry Agent Service is now generally available, Nemotron models are coming to Foundry, and Azure is already powering on NVIDIA Vera Rubin NVL72 in Microsoft labs.
Microsoft says Fireworks AI is now part of Microsoft Foundry, bringing high-performance, low-latency open-model inference to Azure. The launch emphasizes day-zero access to leading open models, custom-model deployment, and enterprise controls in one place.
Azure says GPT-5.4 is now available in Microsoft Foundry for production-grade agent workloads. Microsoft’s supporting post adds GPT-5.4 Pro, pricing, and initial deployment options, with governance controls positioned as part of the pitch.
Microsoft Azure announced that Microsoft Foundry now offers GPT-Realtime-1.5, GPT-Audio-1.5, and GPT-5.3-Codex. The stated focus is low-latency voice interactions and long-running engineering workflows.
Azure posted on February 25, 2026 that three new Azure OpenAI models are rolling out in Microsoft Foundry. Microsoft positions the release for low-latency voice systems and long-running engineering workflows with published pricing and performance claims.