Microsoft Agent Framework
Microsoft's unified SDK for agents and multi-agent workflows
Microsoft Agent Framework is profiled here as a LLM tool for engineering teams. Read about features, pricing, and how it compares to related options in the tools directory.
Description
Microsoft Agent Framework is an open-source SDK and runtime that consolidates Microsoft's two earlier agent projects, Semantic Kernel and AutoGen, into one toolchain. Released in public preview in October 2025, it combines AutoGen's multi-agent orchestration with Semantic Kernel's enterprise features such as state management, telemetry, and type safety. The framework supports Python and .NET, adds explicit graph workflows for control over execution paths, and integrates with Azure AI Foundry. General availability is targeted for early 2026, and Microsoft positions the framework as the successor path for both predecessor projects. Migration guides help teams move existing Semantic Kernel and AutoGen code onto the unified runtime.
Key Capabilities:
Single- and multi-agent orchestration in one API
Graph-based workflows for explicit multi-agent control
MCP and Agent-to-Agent protocol support
Durable state management for long-running and human-in-the-loop runs
Built-in observability, evaluation, and hosting
Python and .NET support with Azure AI Foundry integration
Alternative tools
- Browser Use
Connect AI agents to the browser for web tasks
- OpenAI Agents SDK
Lightweight framework for building multi-agent workflows
- Nomic
Open embedding models with large-scale data visualization
- txtai
All-in-one embeddings database for semantic search and RAG
- Perplexity Sonar API
Search-grounded language model API with live citations
- AI21 Labs
Hybrid Mamba-Transformer models for enterprise applications