Home / Tools / Semantic Kernel

Semantic Kernel

Microsoft SDK for AI agents in your code

Semantic Kernel preview
learn.microsoft.com Open Source

Upvotes

0

Pricing

Open Source

Category

Agents

In the database since

Oct 2026

Overview

About Semantic Kernel

Semantic Kernel is Microsoft's open-source kit for adding AI agents and language models to real software. It works with C#, Python, and Java codebases, so existing apps can gain AI features without a rewrite. Developers combine agents, plugins, planners, and memory into production-ready assistants. It connects to any LLM provider, keeping you free to choose models. The code is MIT-licensed with nearly 29,000 GitHub stars. Microsoft's own docs host the full guides. Prompt templates keep prompts versioned like normal code.

Key Features

1. Three-language SDK

Build in C#, Python, or Java with the same concepts. Teams keep their existing stack and language.

2. Agents and plugins

Agents do the reasoning while plugins expose your code as tools. Your app's functions become actions the AI can take.

3. Planners

Planners break a goal into ordered steps automatically. Multi-step tasks run without hand-written orchestration.

4. Memory and connectors

Built-in memory stores facts across sessions. Connectors link to vector databases and external services.

5. Any model provider

Use OpenAI, Azure, or other models through a common interface. Switching providers rarely means rewriting code.

6. Prompt templates

Manage prompts as versioned templates with variables. Prompt engineering becomes a normal part of the codebase.

Frequently Asked Questions (FAQ)

What does Semantic Kernel cost?

Nothing. The MIT license covers all the code, free forever. You only pay your chosen LLM provider for API usage.

Which languages are supported?

C#, Python, and Java have first-class SDKs. Each follows the same agent, plugin, and planner concepts.

How is this different from LangChain?

Semantic Kernel leans enterprise: strong typing, .NET roots, and Microsoft support. LangChain leans toward rapid Python prototyping.

Where are the docs?

Microsoft Learn hosts the official guides and API references. Source code and samples live on GitHub.

Conclusion

Semantic Kernel is the steady enterprise choice for putting agents inside existing software. The three-language support and Microsoft backing matter to IT departments. It will not dazzle like a demo product, but it ships. If your codebase is .NET or Java, start here before looking elsewhere. The samples repo speeds up first builds nicely.

Newsletter

Tools like this change fast

New Agents tools launch every week — the newsletter keeps you ahead of what's worth trying.

Keep exploring

Similar AI tools

Ranked by the job they do, their capabilities, and how closely they fit this workflow.