Microsoft Agent Framework is Microsoft's open-source framework for building AI agents and multi-agent workflows in Python and .NET, succeeding Semantic Kernel and AutoGen.
In plain words
The framework has a few main parts. Agents use a language model to handle input, call tools and MCP servers, and respond. Workflows link agents and functions into explicit, graph-shaped steps for multi-agent jobs. Integrations connect to model providers such as Microsoft Foundry, Azure OpenAI, OpenAI and Anthropic. It merges AutoGen's simple agent design with Semantic Kernel's enterprise features like state, middleware and telemetry.
A quick example
Imagine a support workflow: one agent reads an incoming ticket, a second searches the knowledge base through a tool, and a third drafts the reply. Agent Framework's workflows connect them in a fixed order you can trace and debug.
Where you'll meet it
- Production agents in Python and .NET.
- Teams moving on from Semantic Kernel or AutoGen.
Related
Sources
- DocsMicrosoft Agent Framework README, Microsoft (GitHub) · read 29 Sept 2026
- DocsMicrosoft Agent Framework overview, Microsoft Learn · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.