Orchard: An open framework for scalable agentic AI
Microsoft open-sources Orchard: a framework for training and evaluating AI agents at scale, lowering the bar for smaller models.
Why it matters
Orchard democratizes agent development by providing reusable infrastructure for training and evaluation across task types. This shifts agent-building from labs with unlimited compute to the research community, potentially accelerating agent-capability diversity.
The key facts
10 to knowOpen-source framework from Microsoft Research
Designed for training and evaluating AI agents across multiple task types
Enables strong performance from smaller models via infrastructure reuse
Reduces complexity for agent research
Targets research community adoption
Open-source framework for training and evaluating AI agents
Designed to enable strong performance from smaller models
Reduces infrastructure complexity for agent research
Supports task-agnostic agent development
Published by Microsoft Research
Go to the source
Microsoft Researchmicrosoft.com
Publisher excerpt: Orchard is an open-source framework for the research community to train and evaluate AI agents across task types. It reduces complexity while supporting strong performance from smaller models by enabling researchers to reuse the same infrastructure.