Updated October 2026
Open source LLMs are language models whose weights you can download and run yourself, together with the code that loads and trains them. Running them locally keeps your data private, avoids per-token fees, and lets you fine-tune a model for your own task.
This page lists the most-starred models and the tooling around them: runtimes such as Ollama and llama.cpp, serving engines such as vLLM, fine-tuning kits, and chat interfaces.
Ranked by GitHub stars and refreshed daily.
by ollama
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other models.
by huggingface
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
by open-webui
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
by ggml-org
LLM inference in C/C++
by deepseek-ai
by vllm-project
A high-throughput and memory-efficient inference and serving engine for LLMs
by unslothai
Local UI to run and train LLMs and diffusion models. Supports GGUF, MLX, Qwen3.8, DeepSeek-V4, MiniMax-H3, Gemma 4, FLUX and more.
by nomic-ai
GPT4All: Run Local LLMs on Any Device. Open-source and available for commercial use.
by hiyouga
Unified Efficient Fine-Tuning of 100+ LLMs & VLMs (ACL 2024)
by karpathy
The simplest, fastest repository for training/finetuning medium-sized GPTs.
by meta-llama
Inference code for Llama models
by mudler
LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
by exo-explore
Run frontier AI locally.
by lyogavin
AirLLM 2.8T LLM inference with single 4GB GPU
by lm-sys
An open platform for training, serving, and evaluating large language models. Release repo for Vicuna and Chatbot Arena.
by sgl-project
SGLang is a high-performance serving framework for large language models and multimodal models.
by QwenLM
Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.
by haotian-liu
[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
by QwenLM
The official repo of Qwen (通义千问) chat & pretrained large language model proposed by Alibaba Cloud.
by AI4Finance-Foundation
FinGPT: Open-Source Financial Large Language Models! Revolutionize 🔥 We release the trained model on HuggingFace.
By GitHub stars, the most popular projects on this list right now are ollama, transformers, and Open WebUI. The ranking is rebuilt every day from GitHub.
The quickest way is a runtime such as Ollama, llama.cpp, or GPT4All: install it, download a model, and chat from the terminal or a local API. For serving many users at once, engines such as vLLM and SGLang are built for throughput.
Not always. Many models publish their weights but not their training data, and some licenses limit commercial use. Check each model's license before you ship it in a product.
