https://github.com/hathibelagal-dev/desktop4mistral
A desktop client with MCP support for Mistral LLMs
https://github.com/hathibelagal-dev/desktop4mistral
artificial-intelligence chat desktop-app gui llm mcp mistral mistralai qt6
Last synced: 2 months ago
JSON representation
A desktop client with MCP support for Mistral LLMs
- Host: GitHub
- URL: https://github.com/hathibelagal-dev/desktop4mistral
- Owner: hathibelagal-dev
- License: gpl-3.0
- Created: 2025-03-10T11:33:21.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-03-26T02:58:46.000Z (over 1 year ago)
- Last Synced: 2025-03-26T03:03:19.046Z (over 1 year ago)
- Topics: artificial-intelligence, chat, desktop-app, gui, llm, mcp, mistral, mistralai, qt6
- Language: Python
- Homepage:
- Size: 1.09 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# desktop4mistral
https://github.com/user-attachments/assets/0ec0bec8-2b35-4c74-9d22-2739935db5b5
A powerful desktop client for interacting with Mistral Large Language Models (LLMs)
[](https://www.gnu.org/licenses/gpl-3.0)
[](https://pypi.org/project/desktop4mistral/)
[](https://pypi.org/project/desktop4mistral/)
## Overview
desktop4mistral is a Python-based desktop application that provides a user-friendly interface for chatting with Mistral AI models. Built with PySide6, it offers a modern GUI with features like model selection, chat history, and command support.
## Features
- Interactive chat interface with Mistral LLMs
- Support for multiple Mistral models with easy switching
- Full Markdown support.
- Command system (e.g., `/read` to fetch any local file or webpage, `wiki_search` to search Wikipedia, etc).
- Some commands also support a more natural language syntax. You can, for instance, say "read the contents of /tmp/myfile.txt".
- Use `/save_markdown` to save your entire chat as a markdown file, which you could use in other tools, like Obsidian.
- Supports Python code execution. Ideally, you should first ask it to write some Python code. In the next prompt you can just say something like "run it". This way you can be sure what the model's doing.
## Commands
Desktop4Mistral supports several commands.
- `/read` to read a local or remote file. Can also be used to reload a previous chat session.
- `/git` to read a github repository
- `/wiki_search` to search Wikipedia
- `/wiki_id` to look up the contents of a Wikipedia page
- `/save` to save the entire chat session as a JSON file
- `/save_markdown` to save the entire chat session as a markdown file
- `/talk` to turn talking `on` or `off`. Uses Kokoro as the TTS model. You can expect reasonable performance on most hardware.
## Screenshots
## Installation
### Prerequisites
- Python 3.11 or 3.12
- Mistral API key (get it from [Mistral AI](https://mistral.ai/))
## Quickstart
Install using `pip`.
```bash
pip install desktop4mistral
```
And run...
```bash
export MISTRAL_API_KEY='your-api-key-here'
desktop4mistral
```
or
```bash
python3 -m desktop4mistral.main
```
### Setup for development
1. Clone the repository:
```bash
git clone https://github.com/hathibelagal-dev/desktop4mistral.git
cd desktop4mistral
```
2. Install the app and its dependencies:
```bash
pip3 install .
```
## Usage
- Launch the application
- Select a Mistral model from the "Models" menu
- Type your message in the input field
- Press Ctrl+Enter or click "Send" to submit
- View responses in the chat window
## Support
For issues and feature requests, please use the GitHub [Issues](https://github.com/hathibelagal-dev/desktop4mistral/issues) page.
## License
This project is licensed under the GNU General Public License v3 (GPLv3) - see the LICENSE file for details.