Semantic Kernel

Semantic Kernel is Microsoft's open-source SDK for orchestrating LLMs through plugins, planners, and agents, with strong C#/.NET support alongside Python and Java. It is free and self-hostable and well-suited to enterprise Microsoft environments, and as of 2026 it is converging with AutoGen into the unified Microsoft Agent Framework.

Semantic Kernel (SK) approaches AI as plugin orchestration: you expose native functions and prompt functions as callable skills, and the kernel coordinates them, adds memory and connectors, and lets models invoke them as tools. It has been especially popular in enterprise .NET environments because it offers first-class C# support, type safety, dependency injection, and telemetry that fit existing Microsoft stacks, while also supporting Python and Java. As with AutoGen, the 2026 story is consolidation. Microsoft has merged the lineages of Semantic Kernel and AutoGen into the Microsoft Agent Framework, which reached 1.0 in April 2026 and is presented as the production successor for new agent development. Semantic Kernel contributed much of the enterprise scaffolding, such as session-based state management, middleware, and telemetry, to that framework. Microsoft has signaled a supported migration path rather than an abrupt end, so existing SK applications are not stranded. Semantic Kernel remains a solid, free, open-source choice for adding AI capabilities to enterprise applications, particularly in .NET. But teams starting new agent-heavy projects should understand that the strategic direction now runs through the Microsoft Agent Framework, and plan accordingly.