Microsoft is rolling out OpenAI’s new free and open GPT model, gpt-oss-20b, to Windows 11 users through its Windows AI Foundry. The platform lets users run AI features and open-source models right on their PCs.
OpenAI’s gpt-oss-20b launched Tuesday and can run on consumer hardware with at least 16GB VRAM, typically found in recent Nvidia or Radeon GPUs. It’s designed for agentic tasks like code execution and tool use, running efficiently even on bandwidth-limited setups.
The model is text-only and won’t handle images or audio. It’s been trained with high-compute reinforcement learning to boost performance on AI agent tasks like web search and Python code execution. 
The catch: gpt-oss-20b hallucinates heavily, misfiring on 53% of questions in OpenAI’s PersonQA test, which measures factual knowledge.
Microsoft plans to bring gpt-oss-20b to macOS and other devices soon but hasn’t named specifics. Both gpt-oss-20b and the older, bigger gpt-oss-120b are also available through Azure AI Foundry. Amazon’s AWS has both models too.
Microsoft called gpt-oss-20b:
“tool-savvy and lightweight”
“Optimized for agentic tasks like code execution and tool use, it runs efficiently on a range of Windows hardware, with support for more devices coming soon. It’s perfect for building autonomous assistants or embedding AI into real-world workflows, even in bandwidth-constrained environments.”
