Microsoft Agent Framework review

Microsoft Agent Framework is the direct successor to AutoGen and Semantic Kernel, built by the same teams. It keeps AutoGen's simple single and multi-agent abstractions and adds Semantic Kernel's enterprise pieces: session-based state, type safety, middleware, filters and telemetry. The framework has four parts. Agents call tools and MCP servers on Microsoft Foundry, Anthropic, Azure OpenAI, OpenAI, Ollama and other providers. The Harness Agent adds planning, todo tracking, context compaction, file access, memory and tool approval for long tasks. Workflows connect agents and functions through functional or graph-based paths, and integrations cover model providers, evaluation services and UI frameworks. It is MIT licensed and supports .NET and Python, with Go in public preview without declarative agents or workflows. Microsoft's own advice is to use a workflow when the steps are known and an agent only when the task is open-ended. Examples assume Azure credentials, so non-Azure teams have more setup.

Who it is for

.NET shops and Azure customers that want enterprise features such as typed state, middleware and telemetry.

Price

Free and open source (MIT); you pay for models and Azure services.

Microsoft Agent Framework (MAF) is an open, multi-language framework for building production-grade AI agents and multi-agent workflows in .NET and Python
Quoted from the vendor pricing page on November 6, 2026.

Key features

• Agents that call tools and MCP servers across several model providers
• Harness Agent with planning, todo tracking, context compaction and tool approval
• Functional and graph-based workflows with explicit execution paths
• Session state, context providers for memory and middleware
• Integrations for evaluation services and UI frameworks

Strengths and weaknesses

• Strength: First-class .NET support, rare among agent frameworks
• Strength: Combines AutoGen's abstractions with Semantic Kernel's enterprise features
• Strength: Built-in tool approval in the Harness Agent
• Weakness: Go version is in public preview with missing features
• Weakness: Quickstarts lean on Azure and Microsoft Foundry
• Weakness: Teams on AutoGen or Semantic Kernel face a migration

Key facts

• License: MIT
• Languages: .NET, Python, Go in public preview
• Replaces: AutoGen and Semantic Kernel
• Install (Python): pip install agent-framework

Where it ranks

Microsoft Agent Framework is No. 6 in our Best AI Agent Frameworks verdict.