PyGPT is a cross-platform desktop AI assistant in Python that integrates with OpenAI and other models, featuring multiple modes, plugins, memory, and file/web capabilities. It provides binaries, PyPi installation, and source-based setup, with ongoing releases and multiple model integrations.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
PyGPT is a desktop AI assistant that interfaces with OpenAI models (including GPT-5 and GPT-4) and other providers via various SDKs, LlamaIndex, and Ollama. It supports chat, file-based interactions, agents, vision, audio, image and video generation, web search, memory, presets, and plugin-based extensions. It runs on Linux, Windows, and Mac, and is available as binaries, via PyPi, or from source.
How it works
The project exports a desktop application built in Python, leveraging multiple models and providers through OpenAI API and alternative LLMs (LlamaIndex, Ollama, HuggingFace, xAI Grok, DeepSeek, Bielik, Gemini, Claude, Perplexity, etc.). It includes features such as memory (short and long-term), presets for prompts, notepad, and tools/commands execution via plugins. It supports audio (speech synthesis/recognition) and image/video generation, as well as file I/O and web search. The README lists numerous capabilities and integrations, but does not provide a single unified runtime diagram.
Getting started
Installation options are provided:
- Binaries (Linux, Windows): download from the project download page; 64-bit binaries only; Linux GLIBC requirement: 2.35.
- Microsoft Store (Windows) installation link.
- AppImage (Linux) with execution permission note and AppImageUpdate tip.
- Snap Store (Linux) installation and update commands, plus device connections for camera/mic/audio.
- PyPi (pip): virtualenv setup, then pip install pygpt-net, then run pygpt.
- Running from GitHub source: steps for pip install or Poetry, environment setup, running run.py, and notes on PyInstaller for packaging.
Commands quoted exactly as written in README include:
sudo snap install pygpt
python3 -m venv venv
source venv/bin/activate
pip install pygpt-net
pygpt
git clone https://github.com/szczyglis-dev/py-gpt.git
cd py-gpt
pip install -r requirements.txt
python3 run.py
poetry env use python3.10
poetry install
poetry run python3 run.py
chmod +x ./PyGPT-X.X.X-x86_64.AppImage
appimageupdatetool ./PyGPT-X.X.X-x86_64.AppImage
sudo snap connect pygpt:camera
sudo snap connect pygpt:audio-record :audio-record
sudo snap connect pygpt:audio-playback
sudo snap connect pygpt:alsa
sudo snap connect pygpt:docker-executables docker:docker-executables
sudo snap connect pygpt:docker docker:docker-daemon
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python3 run.py
poetry shell
poetry env use python3.10
poetry install
poetry run python3 run.py
python3 -m venv venv
source venv/bin/activate
python3 run.py
qt.qpa.plugin: Could not load the Qt platform plugin "xcb" in "" ...
sudo apt install libxcb-cursor0
sudo apt install portaudio19-dev
sudo apt install libasound2
sudo apt install libasound2-data
sudo apt install libasound2-plugins
sudo apt install libxcb-cursor0
python3 -m venv venv
pygpt
python3 -m venv venv
source venv/bin/activate
python3
pygpt
python3 run.py --legacy=1
python3 run.py --disable-gpu=1
{"render.engine": "legacy",
"render.open_gl": false}
Recent releases
- v2.7.12 (2026-02-06): xAI SDK upgraded to v1.6.1; Added video generation support in xAI provider; Added new models grok-imagine-image and grok-imagine-video; UI improvements.
- v2.7.11 (2026-02-05): Improved handling focus between columns; Fixed models debugger update.
- v2.7.10 (2026-02-03): Fixed avatar overwrite when creating a new preset; Fixed new context creation when opening a new tab in the second column.
- v2.7.9 (2026-01-08): Improved realtime audio mode; Added xAI provider and Grok support in realtime audio mode.
- v2.7.8 (2026-01-06): Added xAI Collections remote tool and integrated collections management into Remote Vector Stores tool; Unified the Remote Vector Stores tool into a single tool for all providers.
Traction
Not provided in the README FACTS.
Behind the repo
Not provided in the FACTS.
Caveats
- License: none listed.
- Created: 2023-04-09; last_push: 2026-02-06.
- Open issues: 63.
- Languages: Python.
- Requires an API key for OpenAI models; local models do not require API keys.
- Hardware and OS requirements vary by installation method (GLIBC, platform availability).