Local AI

Ollama

93

Run open-source LLMs locally with one command

Ollama makes it trivial to download and run open-weight language models on your own machine, exposing them through a clean local API that mirrors common SDK shapes. It is the default starting point for developers, privacy-conscious users, and anyone working offline with models like Llama and Qwen.

LM Studio

87

A desktop app for running and chatting with local LLMs

LM Studio is a polished desktop application for discovering, downloading, and chatting with local language models, with built-in model server mode for development. It is popular with engineers who want a friendly UI on top of llama.cpp without giving up the privacy of fully local inference.

AnythingLLM

50

The ultimate all-in-one local AI chat app for desktop and enterprise

The ultimate all-in-one local AI chat app for desktop and enterprise — A desktop application that lets you run local models with offline RAG (Retrieval-Augmented Generation) capabilities.

Fooocus

50

An image generating software based on Stable Diffusion, offline and open source.

An image generating software based on Stable Diffusion, offline and open source. — A powerful local and open-source option for offline image generation.

GPT4All

50

A free-to-use, locally running companion that does not require internet.

A free-to-use, locally running companion that does not require internet. — GPT4All is an open-source ecosystem that enables running quantized models locally on consumer-grade hardware with a GUI.

Jan

50

An open-source alternative to ChatGPT that runs 100% offline on your computer.

An open-source alternative to ChatGPT that runs 100% offline on your computer. — Jan is an open-source local LLM runner with a slick user interface that works natively and entirely offline.

Llamafile

50

Distribute and run large language models with a single file.

Distribute and run large language models with a single file. — Mozilla's Llamafile lets you run LLMs using single-file executables that work across multiple operating systems out of the box.

llama.cpp

50

Port of Facebook's LLaMA model in C/C++ for local inference

Port of Facebook's LLaMA model in C/C++ for local inference — It is the foundational engine that powers Ollama and most modern local LLM runners under the hood.