OpenAI has released a new lightweight open GPT model, and Microsoft is making it accessible for Windows users. The gpt-oss-20b model, designed for PC use, is now available on Windows AI Foundry and will soon be released for macOS as part of Microsoft’s local foundry efforts.
To run this model, users will need a PC or laptop equipped with at least 16GB of VRAM, requiring high-end GPUs from Nvidia or certain Radeon models. The gpt-oss-20b model is optimized for code execution and tool usage, making it ideal for developing autonomous assistants or integrating AI into real-world workflows, even in bandwidth-limited environments.
ICYMI: Cleary confirmed for 2 more years as Cork senior football boss
Microsoft has pre-optimized gpt-oss-20b for local inference and hinted at the possibility of support for additional devices in the near future. This could mean a more tailored version for Copilot Plus PCs may be introduced, following the trend of integrating various local AI models into Windows.
The rapid integration of OpenAI’s latest model into Windows AI Foundry coincides with Amazon’s swift adoption of the new open-weight GPT-OSS models for its cloud services. This marks the first occasion that users can run an OpenAI model locally on Windows, while also highlighting the evolving dynamics of the partnership between OpenAI and Microsoft, especially as Microsoft’s major cloud competitor gains access to the latest models.
SOURCE: THE VERGE
