GPT-4/GPT-3.5 Desktop AI Assistant

Open source personal AI Assistant for Linux, Windows and Mac, featuring chat, vision, completion, agents, image generation, command execution, voice control and more. It operates on the system desktop, utilizing OpenAI API key.

64-bit, release: 2.2.19 (2024-05-16), Changelog, GitHub, PyPi, Snap Store
pip install pygpt-net
sudo snap install pygpt
PyGPT is a third-party software, not associated with OpenAI.
GitHub GitHub Repo stars GitHub Repo stars
MIT License.

Multiple features

Supports GPT-4, GPT-4V, GPT-3.5 and DALL-E 3, as well as all LLMs provided by Langchain. It features built-in internet search, speech synthesis and recognition, presets for prompts, conversation history, plugin support and more.

9 modes of operation

The assistant offers multiple modes of operation: chat, completion, real-time vision analysis, image generation (via DALL-E 3), assistants (via the OpenAI Assistants API), experts, autonomous agents and integration with LLMs provided by Langchain.

Full integration

The assistant can execute system commands, custom user commands, generate and execute Python code, and connect with any service, including access to the Web. It allows the use of both local and external tools.

Open Source

The application is free to use. The only requirement is your own OpenAI API key. The source code can be found on GitHub. Compiled binary versions are available for both Windows and Linux.


  • Desktop AI Assistant for Linux, Windows, and Mac.
  • Supports multiple models: GPT-4, GPT-4 Vision, GPT-3.5 and any Langchain LLMs.
  • 9 modes of operation: chat, assistant, completion, image generation, autonomous agents, experts, vision analysis, Langchain and Chat with files (Llama-index).
  • Included support features for individuals with disabilities: customizable keyboard shortcuts, voice control, and translation of on-screen actions into audio via speech synthesis.
  • Speech synthesis using Microsoft Azure, OpenAI TTS, Google and Eleven Labs.
  • Speech recognition using OpenAI Whisper, Google and Microsoft Bing.
  • Real-time internet access via Google and Microsoft Bing.
  • Execution of Python code and system commands.
  • Integration with OS and external services via API calls.
  • Access to the OS filesystem.
  • Vector embeddings and chat with files via Llama-index.
  • Image generation using DALL-E 3.
  • Image analysis using GPT-4 Vision.
  • Short and long-term memory capabilities.
  • Storage and restoration of context and conversation history.
  • Editable prompt presets.
  • Ability to support future OpenAI models, like GPT-5.
  • Extendable with plugins.
  • Built-in token usage calculation.
  • Open source: full source code available on GitHub.
  • Uses the user’s API key.

Download for free
or check it out on GitHub.

Download View source code on GitHub

Download v.2.2.19

build 2024-05-16, 64-bit

install with PyPi:

pip install pygpt-net

install with Snap:

sudo snap install pygpt

Get it from the Snap Store
This website uses cookies to store session information and gather data for analytics. By using our site, you agree to our use of cookies.