GPT4All
Nomic's free offline LLM app, no longer actively developed
Overview
A free, MIT-licensed desktop app from Nomic AI that runs open-weight LLMs offline, with LocalDocs for private document chat. There has been no release since February 2025, and Nomic says it is moving the app to maintenance mode.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 12 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
What you can produce with GPT4All
- Open-weight GGUF models run fully offline on Windows, macOS and Linux
- LocalDocs private chat over your own files
- GPU acceleration through Vulkan (including AMD), Metal and CUDA
- An OpenAI-compatible local API server with LocalDocs access
- A built-in JavaScript code interpreter (since v3.6, December 2024)
- Optional remote models from Groq, OpenAI and Mistral
- MIT licensing, free for commercial use, with no updates since February 2025
GPT4All made local LLMs approachable, and it still installs and runs: pull a GGUF model, point LocalDocs at a folder and chat offline with nothing leaving the machine, with GPU acceleration through Vulkan, Metal and CUDA.
But the last release was v3.10.0 in February 2025, and in December 2025 Nomic's chief executive said the company was moving the app to "non-maintained maintenance mode" while Nomic pivoted to software for construction firms. Newer model architectures will not load and no fixes are coming. Fine for an existing user on older models; anyone starting today should choose Jan or LM Studio.
What we found
GPT4All is the free desktop app Nomic AI released in 2023 to run open-weight language models on a personal computer, with no account and no cloud. It is MIT-licensed, installs on Windows, macOS and Linux, and downloads GGUF models from inside the app. LocalDocs, its private document chat, indexes a folder of your files and answers from them without anything leaving the machine.
The feature list is real. GPU acceleration runs through Vulkan, which Nomic launched in September 2023 and which covers AMD as well as NVIDIA cards, alongside Metal on Macs and a CUDA backend. A local server speaks the OpenAI API format and can draw on LocalDocs. Version 3.6.0, in December 2024, added a built-in JavaScript code interpreter, and version 3.8.0 brought support for DeepSeek-R1 distilled models. Earlier versions of this page presented those as 2026 additions.
The problem is what has happened since. The last release, 3.10.0, shipped in February 2025, and the repository's last commit was a build-configuration change in May 2025. In December 2025, answering a user who asked whether the project was dead, Nomic founder and chief executive Andriy Mulyar wrote that the company was "working to formally put the app in non-maintained maintenance mode". Nomic's own site now sells an AI platform for architecture, engineering and construction firms. The GPT4All page still serves installers, but nothing on it is for sale, and the enterprise edition page Nomic used to run now returns a 404.
That changes the advice. The app still works, and for a fixed set of older models it remains a simple way to chat with documents offline. But open GitHub issues report that newer model architectures do not load, and without releases no fixes are arriving. The usage figures sometimes quoted for it, 250,000 monthly users and 70,000 monthly Python downloads, are Nomic's own numbers from July 2024. Newcomers should look at Jan or LM Studio, which still ship releases.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including GPT4All's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with GPT4All unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to GPT4All directly →
Spotted something out of date? Suggest an update →
Alternatives to GPT4All
More in LLMs & Chat
