Buying snapshot
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Decision point | What Microsoft states |
|---|---|
Languages and runtimes | Python 3.10+, .NET 10.0+ and Java (JDK 17+) |
Operating systems | Windows, macOS and Linux |
Model connections | Built-in support for OpenAI, Azure OpenAI, Hugging Face, NVidia and more; local models through Ollama, LMStudio or ONNX |
Help and community | GitHub issues and the Discord community for problems; regular office hours and community events |
Our take
If you already have agents or plugins built on Semantic Kernel, it remains a workable MIT-licensed SDK to keep running while you plan your next step. If you are starting a new agent project, evaluate Microsoft Agent Framework first, because Microsoft presents it as the successor to Semantic Kernel.
Evaluate Semantic Kernel
Comparisons
Where it stands out
Several ways to give agents tools. Plugins can come from native code functions, prompt templates, OpenAPI specs or the Model Context Protocol (MCP), so existing business code and APIs can become agent tools.
Structured workflows, not only chat. The Process Framework models complex business processes with a structured workflow approach, and multi-agent support lets specialist agents collaborate. The README's example has a triage agent that forwards requests to billing and refund agents.
Enterprise components in the docs. The documentation groups Observability, Security and Filters as enterprise components.
Retrieval and multimodal inputs. It integrates with vector stores such as Azure AI Search, Elasticsearch and Chroma, documents Memory as a core concept, and processes text, vision and audio inputs.
Plenty of starting code. The project points to over 100 detailed samples, plus C# and Python API references.
Product screenshot
Trade-offs to weigh
Microsoft points new work to a successor. The repository says Microsoft Agent Framework is the enterprise-ready successor to Semantic Kernel and links a Semantic Kernel migration guide. Code you write on Semantic Kernel today may need that migration later.
Two things called Agent Framework. The Semantic Kernel documentation has its own Agent Framework page under Frameworks, while the successor product is named Microsoft Agent Framework. Check which one a tutorial means before you follow it.
Microsoft steers the core. For new features, components or extensions, the project asks contributors to open an issue before sending a pull request, because the team "might be taking the core in a different direction". Plan extensions you need on your side rather than expecting them upstream.
Pricing and plan conditions
Semantic Kernel is a package you install into your own application, so the bill does not come from the SDK. Usage cost comes from the AI service you connect it to.
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Item | What the project states |
|---|---|
License | MIT license, copyright Microsoft Corporation |
Installation | Python: pip install semantic-kernel. .NET: dotnet add package Microsoft.SemanticKernel and Microsoft.SemanticKernel.Agents.Core |
AI service access | Set AZURE_OPENAI_API_KEY for Azure OpenAI, or OPENAI_API_KEY for OpenAI directly |
Public signals
About these signals
Verified factual signals from public sources. They indicate observable activity or interest, not total adoption, product quality, or cost.
68 GitHub commits 90d28.7k GitHub stars0 vulnerabilities across 1 package
See all signals from 6 sources
PyPI · semantic-kernel@1.45.0
Alternatives
Microsoft Agent Framework. Microsoft describes version 1.0 as a production-ready release with stable APIs and a commitment to long-term support. It offers multi-agent orchestration, multi-provider model support, and cross-runtime interoperability via A2A and MCP.
Alternatives to Semantic Kernel
The reviewed substitutes for Semantic Kernel among the agent frameworks, and what would make each one the better answer.
Direct alternatives
Reviewed substitutes: products bought for the same job, where a team picks one.
- Agno
- Developers who want to add LLM agents to their code can start with Agno or with Semantic Kernel, since both are open-source frameworks for agents with tools, memory and any model they choose. Semantic Kernel is Microsoft's SDK with a plugin model for C#, Python and Java, while Agno is a Python framework that adds its own AgentOS runtime and control plane for agents in production.Applies to: Choosing between these two for the ai agent frameworks decision.
- AutoGen
- AutoGen and Semantic Kernel are open-source Microsoft projects for putting LLM agents into applications. AutoGen is a Python framework built around multi-agent conversation, with a no-code Studio for trying ideas, while Semantic Kernel is an SDK for C#, Python and Java that wires models into existing code through plugins and planners.Applies to: Choosing between these two for the ai agent frameworks decision.
- Haystack
- Semantic Kernel and Haystack are open-source frameworks for building LLM applications with agents, tools and retrieval. Semantic Kernel, from Microsoft, suits teams working in .NET or Java as well as Python and organises features as plugins, while Haystack is a Python framework that builds agents and RAG systems from modular pipeline components.Applies to: Choosing between these two for the ai agent frameworks decision.
- MetaGPT
- MetaGPT and Semantic Kernel are open-source frameworks for adding LLM agents to software. MetaGPT is a Python multi-agent framework in which agents take team roles and collaborate on one goal, while Semantic Kernel is Microsoft's SDK for C#, Python and Java that connects models to existing code through plugins and planners.Applies to: Choosing between these two for the ai agent frameworks decision.
Other approaches
A different approach to the same problem. Each substitutes only for the workload named beside it.
- LangGraph
- LangGraph is preferred over Semantic Kernel for stateful agent workflows requiring cycles, persistence, human approval, and explicit control flow. We recommend LangGraph over Semantic Kernel when an analytics or data platform team must model branching execution, approvals, and durable state explicitly, accepting the need to design more of the workflow itself.
- AutoGPT
- AutoGPT is an alternative to Semantic Kernel for digital-workflow assistant workloads where a hosted automation product is the preferred operating model. **AutoGPT** provides intelligent assistants intended to streamline digital workflows, with both self-hosted open-source use and a hosted cloud offering.
- CrewAI
- Semantic Kernel is preferred over CrewAI for application development workloads that need an open-source SDK with planners and plugin architecture. **Semantic Kernel** is Microsoft’s open-source SDK for incorporating LLMs into applications through AI agents, planners, and a plugin architecture.
- Google Agent Development Kit
- Google's open agent framework — hierarchical multi-agent composition, graph workflows that mix deterministic code with model reasoning, a local dev UI and evaluation harness, and a deployment path onto Vertex AI Agent Engine. Python, TypeScript, Go, Java and Kotlin.
- LangChain
- Both build applications where a model calls tools and takes several steps, from different starting points: framework breadth, explicit orchestration, role decomposition or an SDK inside an existing application. Teams pick one to build on, and the choice follows the shape of the workflow rather than a feature count.Applies to: Choosing the framework an agent or multi-step LLM application will be built on.
- LlamaIndex
- Semantic Kernel and LlamaIndex both orchestrate an LLM over enterprise data with memory, connectors and planning. Semantic Kernel is Microsoft's SDK for C#, Python and Java and suits a .NET or Azure-aligned shop, while LlamaIndex is Python-first and adds an ingestion and indexing layer that Semantic Kernel has no equivalent for.Applies to: Building LLM applications over internal data inside a Microsoft stack. Semantic Kernel stands in for LlamaIndex when the codebase is .NET and the deployment is Azure; LlamaIndex stands in when the work is Python-first and the hard part is ingesting and indexing a document corpus.
- OpenAI Agents SDK
- Two vendor-aligned frameworks separated by the estate rather than the feature list: Semantic Kernel for .NET and Azure with enterprise process orchestration, the OpenAI Agents SDK for teams committed to OpenAI models wanting the shortest path to production. Semantic Kernel is superseded: Microsoft merged it and AutoGen into the Microsoft Agent Framework, GA April 2026, and SK v1.x now receives critical bug and security fixes only, with new features built in the successor. It stays a real option because a .NET estate running it today is supported into 2027, and the catalogue records the supersession on its own page -- but the pair is a choice between a live framework and one on a known sunset, and saying so is the point.Applies to: Picking an agent framework that matches the stack already in place. Semantic Kernel stands in for a .NET or Azure shop already running it; the OpenAI Agents SDK stands in for a Python team on OpenAI models. A fresh Microsoft build belongs on the Agent Framework successor.
- PydanticAI
- Named together in the 2026 guidance as the answer to one requirement -- 'you want typed tool interfaces and explicit runtime control: Microsoft Agent Framework or PydanticAI' -- which is a two-name shortlist for a stated need. The condition is the stack: PydanticAI is the idiomatic-Python answer, Semantic Kernel the Microsoft-aligned one. Note that the guidance names the *successor*: Semantic Kernel is superseded: Microsoft merged it and AutoGen into the Microsoft Agent Framework, GA April 2026, and SK v1.x now receives critical bug and security fixes only, with new features built in the successor. It stays a real option because a .NET estate running it today is supported into 2027, and the catalogue records the supersession on its own page -- but the pair is a choice between a live framework and one on a known sunset, and saying so is the point.Applies to: An agent loop with typed tool interfaces, validated structured output and explicit runtime control. PydanticAI stands in for a Python team that wants ordinary Python; Semantic Kernel stands in inside a .NET estate already running it, and a new one should evaluate the Microsoft Agent Framework instead.