GPT4All

GPT4All is the open-source ecosystem for running large language models (LLMs) privately and locally on your everyday desktop or laptop: no cloud, no GPU required.

GPT4All delivers private, high-performance AI directly to your device (Windows, macOS, Linux). Developed by Nomic AI, this open-source project allows you to run quantized LLMs—like Meta-Llama-3-8B-Instruct.Q4_0.gguf (a 4.66GB model)—on consumer-grade CPUs with just 4GB–16GB of RAM. The platform ensures maximum privacy and security: your data never leaves your machine. It supports a full Python SDK and a desktop chat application, empowering developers and power-users to build custom, secure AI workflows with local document chat (LocalDocs).

https://gpt4all.io