Microsoft is taking a practical step forward by integrating OpenAI’s newest GPT model directly into Windows. This update means you can now run the free, open-source gpt-oss-20b model right on your personal computer through the Windows AI Foundry—and soon on macOS too.
To get the most out of this advanced AI, you’ll need a system with at least 16GB of VRAM supported by a high-performance GPU from Nvidia or Radeon. The model is optimised for executing code and leveraging various tools, making it ideal for creating smart digital assistants or embedding AI into your daily workflow—even when bandwidth is limited.
Microsoft’s move to tailor the gpt-oss-20b for local inference hints at even broader future support across different devices. This approach complements initiatives like the Copilot Plus PCs, part of its continuing effort to bring local AI functionality to everyday computing. It’s interesting to note that this integration follows closely on the heels of Amazon adopting the same open-weight GPT-OSS models for its cloud services, adding another layer to how major tech players are reshaping the AI landscape.
