GPT4All is an open-source tool for running large language models locally on various devices without requiring API calls or GPUs. It supports multiple platforms and integrates with popular tools like Langchain and Weaviate.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
GPT4All is an open-source tool designed to run large language models (LLMs) locally on desktops and laptops. It does not require API calls or GPUs, making it accessible for everyday use.
How it works
GPT4All allows users to download and run LLMs directly on their devices. It supports various model architectures and provides a Python client for easy integration. The tool can be used with models such as Meta-Llama-3-8B-Instruct.Q4_0.gguf.
Getting started
To install GPT4All, you can use the following command:
pip install gpt4all
To use it in Python:
from gpt4all import GPT4All
model = GPT4All("Meta-Llama-3-8B-Instruct.Q4_0.gguf") # downloads / loads a 4.66GB LLM
with model.chat_session():
print(model.generate("How can I run LLMs efficiently on my laptop?", max_tokens=1024))
Recent releases
- v3.10.0 (2025-02-25): Added support for remote model providers including Groq, OpenAI, and Mistral.
- v3.9.0 (2025-02-05): Fixed issues with LocalDocs and DeepSeek-R1 reasoning.
- v3.8.0 (2025-01-31): Introduced support for DeepSeek-R1-Distill models.
- v3.7.0 (2025-01-23): Added support for Windows ARM platform.
- v3.6.1 (2024-12-20): Fixed bugs from v3.6.0 related to conversation controls.
Traction
- Stars: 77386
- Forks: 8304
- Open Issues: 768
Behind the repo
The repository is maintained by Nomic AI, and it is made possible by their compute partner, Paperspace.
Caveats
- License: MIT
- Open Issues: 768
- Age: Created on 2023-03-27, with the last push on 2025-05-27.