Open-source Python library from NVIDIA that adds programmable safety, topic and security rails between your app and an LLM.
Open source
Microsoft Agent Framework
Open, multi-language framework from Microsoft for building, orchestrating and running AI agents and multi-agent workflows in Python, .NET and Go.
What is Microsoft Agent Framework?
Microsoft Agent Framework (MAF) is aimed at developers who build production-ready AI agents and multi-agent workflows. You write agents in Python or C#/.NET with consistent APIs, and a separate repository covers a Go SDK. Beyond a single prompt or chat loop, it provides graph-based orchestration with sequential, concurrent, handoff and group collaboration patterns, plus checkpointing, streaming, human-in-the-loop control and time-travel debugging. A middleware system handles request and response processing, OpenTelemetry integration supports tracing and monitoring, and YAML declarative agents make definitions easy to version. Supported providers include OpenAI, Azure OpenAI, Foundry and the GitHub Copilot SDK, and agents can be deployed to Foundry-hosted infrastructure with two extra lines of code. Migration guides cover Semantic Kernel and AutoGen. You install it with pip install agent-framework or the Microsoft.Agents.AI NuGet package. The README lists no price for the framework itself; model and hosting usage is billed by the services you connect.
Use cases
- Build multi-agent workflows
- Orchestrate agents with graph-based patterns
- Add human approval to agent steps
- Trace and monitor agents with OpenTelemetry
- Host agents on Foundry infrastructure
- Migrate from Semantic Kernel or AutoGen
- Define agents declaratively in YAML
Who it's for
Why use Microsoft Agent Framework
- One framework covers prototype through production, with orchestration, durability and observability included.
- Python and .NET teams get consistent APIs, so mixed-language organizations can share patterns.
- Provider flexibility lets you change model services without a large rewrite.
- Human-in-the-loop and checkpointing support suit governed, long-running agent processes.
- Migration guides lower the effort of moving from Semantic Kernel or AutoGen.
Microsoft Agent Framework features
- Python and C#/.NET implementations with consistent APIs
- Go SDK in a separate repository
- Support for several model providers, including Foundry, Azure OpenAI and OpenAI
- Graph-based workflows: sequential, concurrent, handoff and group collaboration
- Checkpointing, streaming, human-in-the-loop and time-travel
- Middleware for request and response processing and exception handling
- Built-in OpenTelemetry integration for tracing and monitoring
- Declarative agents defined in YAML
- Agent Skills built from files, inline code or class libraries
- DevUI for developing, testing and debugging workflows
- Foundry Hosted Agents deployment
- AF Labs experimental packages for benchmarking and reinforcement learning
Microsoft Agent Framework integrations
How to get started with Microsoft Agent Framework
- Install the package with pip install agent-framework, or add Microsoft.Agents.AI through NuGet for .NET.
- Authenticate to your model provider, for example with Azure CLI credentials for Microsoft Foundry.
- Create an agent with a client, a name and instructions, then call run to get a response.
- Compose agents into graph-based workflows and add tools, middleware and checkpoints.
- Enable OpenTelemetry for tracing and deploy locally or to Foundry-hosted infrastructure.
Microsoft Agent Framework limitations
- The Go SDK is maintained in a separate repository
- Experimental modules ship in the separate agent-framework-lab package
- Third-party services you connect are governed by their own terms and costs
- You are responsible for responsible-AI mitigations and usage costs
Microsoft Agent Framework pricing
Microsoft Agent Framework is open source, so you can use and self-host it for free.
Prices can change and may vary by country or exclude tax. Check the official pricing page.
Microsoft Agent Framework alternatives
The main Microsoft Agent Framework alternatives listed on Debuly are LangGraph, CrewAI, OpenAI Agents SDK, AutoGen, Semantic Kernel and Agent Development Kit (ADK). The table compares pricing and platforms; order is A–Z.
| Tool | Pricing | Starting price | Platforms |
|---|---|---|---|
| Microsoft Agent Framework | Open source | — | Python, .NET, Go, Self-hosted |
| Agent Development Kit (ADK) | Open source | — | CLI, Self-hosted, API |
| AutoGen | Open source | — | API, CLI |
| CrewAI | Open source | — | Web, API, CLI, Self-hosted |
| LangGraph | Open source | — | Self-hosted, Web, API |
| OpenAI Agents SDK | Open source | — | Python, macOS, Linux, Windows |
| Semantic Kernel | Open source | — | Python, .NET, Java, Windows |
Agent Development Kit (ADK)
Google's open-source, code-first Python framework for building, testing and deploying AI agents and multi-agent workflows.
AutoGen
Microsoft's MIT-licensed Python and .NET framework for multi-agent AI apps, now archived; new projects are pointed to Microsoft Agent Framework.
Open sourceVisit AutoGen (official site)
CrewAI
Multi-agent framework and platform for building AI agent crews in Python or a visual editor, then deploying and monitoring them.
Open sourceVisit CrewAI (official site)
LangGraph
Low-level open-source orchestration framework and runtime for building stateful, long-running AI agents as graphs.
Open sourceVisit LangGraph (official site)
OpenAI Agents SDK
Open-source Python framework for building multi-agent workflows with tools, handoffs, guardrails, tracing, sandboxes and voice agents.
Open sourceVisit OpenAI Agents SDK (official site)
Semantic Kernel
Microsoft's open-source SDK for building AI agents and multi-agent systems in Python, .NET and Java, now succeeded by Microsoft Agent Framework.
Open sourceVisit Semantic Kernel (official site)
Microsoft Agent Framework FAQ
How much does Microsoft Agent Framework cost?
The README does not list a price for the framework. You install it from PyPI or NuGet, and you pay for the model, hosting and other services you connect.
Is Microsoft Agent Framework open source?
The project is published openly on GitHub at microsoft/agent-framework and described as an open framework. Check the repository's license file for exact terms.
Which languages does it support?
It supports Python and C#/.NET with consistent APIs. The Go SDK, docs and samples live in microsoft/agent-framework-go.
Which AI providers can I use?
It works with OpenAI, Azure OpenAI, Foundry and the GitHub Copilot SDK, and the README says more providers are on the way.
Can I migrate from Semantic Kernel or AutoGen?
Yes. Microsoft publishes migration guides from Semantic Kernel and from AutoGen in the framework documentation.
What are alternatives to Microsoft Agent Framework?
Alternatives include LangGraph, CrewAI, OpenAI Agents SDK, AutoGen and Semantic Kernel.
Sources & verification
Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.
| Source | Type | Accessed |
|---|---|---|
| github.com/microsoft/agent-framework | official |
Image credits
- Logo: from avatars.githubusercontent.com · fetched · Vendor trademark, used only to identify the product; removed on request
- Cover: none collected yet, so a generated tile is shown.
Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.
More in Developer platforms 4
All developer platforms toolsAsk questions of your data in plain language and get charts and answers, as an open source Python library and a hosted analyst called Annie.
FreemiumOpen-source PostgreSQL extension that adds vector columns, similarity search and HNSW and IVFFlat indexes to your existing database.
Open sourcePay-per-use API and web platform for running 1,000+ image, video, audio and language models without managing GPUs.
Free trial
