Get LM-Kit.NETUp and Running in Minutes.
Install via NuGet, pick a model, and build your first AI-powered .NET application. Full feature access with no time limits, no registration required, and zero cloud dependency.
Install via NuGet
Quick Start Path
From Zero to AI in Three Steps
Create a new project, install the NuGet package, and follow the getting started guide to build your first AI application.
Create a .NET Project
Start with a new console application or add LM-Kit.NET to any existing .NET project.
Install the Package
Add LM-Kit.NET from NuGet. The package includes everything you need for CPU inference.
Build Your First AI App
Follow the getting started guide to initialize the runtime, load a model, and generate your first response.
Your First AI AgentReady-to-Run Code Samples
Clone and run complete .NET applications covering chatbots, RAG pipelines, multi-agent orchestration, document processing, vision, speech recognition, and more. Every sample includes a detailed walkthrough in the developer guide.
Install LM-Kit.NET via NuGet
One command is all you need. LM-Kit.NET is distributed as a single NuGet package that includes CPU inference out of the box with AVX/AVX2 optimization.
Optional: GPU Acceleration
For NVIDIA GPU acceleration, install one additional backend package matching your CUDA version and operating system. Vulkan and Metal support are included in the main package.
Choose the Right Model for Your Use Case
The model catalog is continuously updated with state-of-the-art releases from Google, Meta, Alibaba, Mistral, Microsoft, IBM, and more. It also includes LM-Kit fine-tuned models optimized for specific tasks like sentiment analysis and sarcasm detection. Models auto-download on first use. See how local inference cuts your costs compared to cloud APIs.
var model = LM.LoadFromModelID("gptoss:20b");
var chat = new MultiTurnConversation(model);
var response = chat.Submit("What are the benefits of on-device AI?");
Console.WriteLine(response);
Chat & Code
ChatReasoning & Agents
ReasoningVision & Multimodal
VisionOCR & Document
OCREmbeddings & Reranking
Speech to Text
SpeechWhat You Can Build
A single SDK covering the full spectrum of AI capabilities. All features run 100% on-device with zero cloud dependency. Explore all use cases
AI Agent Orchestration
Document Intelligence
RAG & Knowledge
Vision & Speech
Text & NLP
Text Generation
Chatbots & Assistants
MCP & Tool Integration
Data Extraction
Part of the Microsoft .NET AI Ecosystem
LM-Kit.NET integrates natively with Microsoft Semantic Kernel and Microsoft.Extensions.AI, making it a drop-in local provider for existing .NET AI applications.
Works Everywhere You Build .NET
Targets .NET Standard 2.0 for maximum compatibility. Develop and deploy on your platform of choice.
Operating Systems
.NET Frameworks
Development Tools
Frequently Asked Questions
Common questions about getting started with LM-Kit.NET.
Do I need a GPU to use LM-Kit.NET?
Is the Community Edition really free?
What models are supported?
Does it work offline or in air-gapped environments?
How does local inference compare to cloud APIs?
Can I integrate with Microsoft Semantic Kernel or Extensions.AI?
Ready to Deploy? Get the Free Community License
The Community Edition gives developers, startups, and open-source projects full, unrestricted access to LM-Kit.NET at no cost. No time limits. No feature locks. No strings attached.