#101101st of 153 in Agent frameworks and SDKs
llamasharp
C# library for running GGUF models locally via llama.cpp bindings
- Stars
- 3.8k
- License
- MIT
- Last commit
- Nov 2026
- Last release
- Oct 2026
- Language
- C#
3,598 on 25 Apr 20263,827 on 14 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
LLamaSharp is a .NET library that wraps llama.cpp so C# applications can load GGUF models and run inference on CPU or GPU. It ships NuGet backend packages for CPU, CUDA 12, CUDA 13 and Vulkan, with Metal on Mac through the CPU package. Higher-level APIs cover chat sessions, executors and RAG, and the repo includes console and ASP.NET examples.
Who it is for: .NET developers embedding local LLM inference in C# apps
Strengths
- Prebuilt NuGet backends for CPU, CUDA 12/13, Vulkan and Metal; no C++ build needed
- Native C# API with streaming chat sessions, executors and GPU layer offload
- Includes LLaVA multimodal examples and RAG support
- MIT license; releases are mapped to specific llama.cpp commits
Weaknesses
- Each release is tied to one llama.cpp commit; mismatched backends or old GGUF files crash
- Only loads GGUF; PyTorch or Hugging Face weights need manual conversion
- Bundled ASP.NET project is a demo, not a documented production server
- Integrations such as BotSharp and LangChain live in separate repositories
What it needs
- GPU optional
- Needs llama.cpp native backend NuGet packages, .NET
- Models: GGUF, LLaMA, Mixtral, Phi-2, LLaVA
What changed
- v0.30.0 ↗ (opens in a new tab)Breaking
“feat(cuda): add CUDA 13.0.0 jobs to the compile workflow by @modelpath-dev in https://github.com/SciSharp/LLamaSharp/pull/1439”
- v0.29.0 ↗ (opens in a new tab)Breaking
“Updated to llama.cpp 815a2a5915f22ce6a760c676389c5dfe8535c08f (b10221) by @martindevans in https://github.com/SciSharp/LLamaSharp/pull/1414”
- v0.27.0 ↗ (opens in a new tab)Breaking
“Windows arm64 cpu support by @jihadkhawaja in https://github.com/SciSharp/LLamaSharp/pull/1226”
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in Agent frameworks and SDKs
See all 153| Rank | Project | Score |
|---|---|---|
| 1 | ppt-masterAgent skill that turns documents into natively editable PowerPoint decks | 96 out of 100 |
| 2 | copilotkitReact and Angular SDK for adding agent chat and generative UI to apps | 92 out of 100 |
| 3 | openai-agents-pythonPython SDK for multi-agent workflows with handoffs, guardrails and tracing | 92 out of 100 |
| #4 | mastraTypeScript framework for building AI agents, workflows and MCP servers | 92 out of 100 |
| #5 | browser-usePython library that lets an LLM agent control a web browser | 90 out of 100 |