GPT4All
Run Local LLMs with GPT4All
A free, local, privacy-aware chatbot. No GPU required — runs on CPU with decent speed for smaller models.
Install GPT4All
Download the installer from gpt4all.io. Available for macOS, Windows, and Linux. GPT4All focuses on CPU inference, making it ideal for machines without a dedicated GPU.
Choose and Download a Model
On first launch, GPT4All prompts you to download a model. It recommends models based on your available RAM. Smaller models like Phi-3 Mini work well on 8 GB RAM machines.
Chat Locally
GPT4All provides a simple chat interface. All processing happens on your machine — no internet connection required after the initial download. Great for privacy-sensitive use cases.
LocalDocs (RAG)
GPT4All can index your local documents and answer questions about them. Point it at a folder, and it builds a local embedding index. Useful for searching personal notes, codebases, or documentation.