Source
Orchard: An open framework for scalable agentic AI
Microsoft introduced Orchard, an open-source framework and environment service designed to make the development and evaluation of autonomous AI agents more scalable and cost-effective.
Why it matters
It reduces reliance on proprietary infrastructure, allowing more researchers to build efficient AI agents that can perform real-world tasks like coding and web browsing.
The details
- Orchard-SWE achieves 69.7% on SWE-bench Verified using ~3 billion active parameters.
- The framework supports software engineering, web-navigation, and personal-assistant AI agents.
- Orchard Env uses Kubernetes to manage isolated components for training and evaluation.
Get the weekly recap
The stories like this one, picked and explained — once a week, straight to your inbox.