LM-Kit.NET is a cross-platform software development kit (SDK) designed for .NET applications, enabling on-device execution of large and small language models without relying on external cloud services. It supports a broad range of natural language processing tasks, including text generation, translation, summarization, and entity recognition, executed locally on CPU or GPU, with compatibility across Windows, Linux, and macOS. The SDK provides native integration through a NuGet package, allowing developers to embed AI capabilities directly into desktop, server, or edge applications while maintaining full control over data, performance, and compliance requirements. It supports multiple model architectures and inference backends, giving developers flexibility to tailor performance and resource usage to their specific environments.
The LM-Kit.NET 2026.3.2 update introduces the SearchHighlightEngine, a new feature that enables applications to search and visually highlight text within paginated documents such as PDFs or images. It supports exact matches, regular expressions, and fuzzy matching using the Damerau–Levenshtein distance, allowing developers to locate text even when small spelling variations or recognition errors are present. The engine produces a highlighted version of the document, making results immediately visible and easier to review in document processing, analysis, or verification scenarios. It can also accept pre-computed PageElement instances for raster PDFs or images, allowing existing layout or text extraction results to be reused and reducing the processing required to add search highlighting capabilities.
To see a full list of what's new in version 2026.3.2, see our release notes.
LM-Kit.NET is licensed per deployed commercial application per year. The license is subscription-based and must be renewed every year to continue using the Software. See our LM-Kit.NET licensing page for full details.
Learn more on our LM-Kit.NET product page.
