ClipAI

ClipAI is a simple but powerful utility to connect your clipboard 📋 directly to a Local LLM 🤖 (Ollama-based) such as Gemma 3, Phi 4, Deepseek-V3, Qwen, Llama 3.x, etc. It is a clipboard viewer and text transformer application built using Python.

It is your daily companion for any writing-related job ✏️📄. Easy peasy.

The PC requirements 🖥️ are mostly related to your preferred LLM, but consider that Gemma 3-1B-it-qat-q4_0 runs smoothly on a potato PC 🥔 (8GB RAM and internal GPU).

Features

• Real-time Clipboard Monitoring: Automatically detect and display clipboard content changes
• Text Transformation: Transform clipboard content using various AI-powered operations
• Markdown Support: output text can be view using markdown styling
• Multiple LLM Models: Support for different language models through Ollama
• Customizable Prompts: Configure different transformation types through prompts.json
• Auto-refresh: Toggle automatic clipboard monitoring
• Format Toggle: Switch between plain text and Markdown formatted views
• Copy Output: Easily copy transformed content back to clipboard

Table of Contents

• Features
• Screenshots
• Requirements
• Usage
• Installation/Running
  • Option 1 (run Python script)
  • Option 2 (run executable file, Win Only)
  • Option 3 (build the executable by yourself and run it)
• UI Elements
• Configuration
• Adding New Prompts
• Building Executables
• Project Structure
• Contributing
• License

Screenshots

Requirements

• Python 3.8+
• Ollama running (see https://ollama.com/ to install Ollama)
• Required Python packages:
  • pyperclip
  • requests
  • pyinstaller (only to build the project by yourself, see below how to do it)

Usage

1. Copy any text to your clipboard
2. Select desired transformation type
3. Choose LLM model
4. Click Send button or press Shift+Enter
5. View transformed output
6. Use right-click to toggle formatting
7. Copy output using the copy button

Installation

Option 1 (run Python script)

1. Clone the repository
2. Create a virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # Linux/Mac
   .venv\Scripts\activate     # Windows

3. Install dependencies:

   pip install -r requirements.txt

4. Configure Ollama and ensure it's running
5. Run the application:

   python run.py

Option 2 (run executable file, Win Only)

1. Download the zip file from last Release
2. Extract the files
3. Configure Ollama and ensure it's running
4. Run the .exe

Option 3 (build the executable by yourself and run it)

1. Follow the instruction here

UI Elements

Buttons

1. Refresh Button ()

   • Manually updates the clipboard content
   • Useful when auto-refresh is disabled

2. Auto-refresh Toggle ()

   • Toggles automatic clipboard monitoring
   • indicates active monitoring
   • indicates manual mode

3. Clear Button ()

   • Clears both input and output text areas
   • Resets the current state

4. Send Button ()

   • Sends content to the selected LLM model
   • Changes to stop icon () during processing
   • Click again to stop the current operation

5. Copy Button ()

   • Copies the output content to clipboard
   • Only active when there is content to copy

Dropdowns

1. Transformation Type

   • Selects the type of transformation to apply
   • Options are loaded from prompts.json
   • Examples might include:
     • Summarization
     • Translation
     • Code explanation
     • Text formatting
     • Custom transformations

2. Model Selection

   • Chooses the LLM model to use
   • Automatically populated from available Ollama models
   • Default model is set in config.json

Text Areas

1. Input Area

   • Displays current clipboard content
   • Updates automatically when auto-refresh is enabled
   • Supports manual updates via refresh button

2. Output Area

   • Shows transformed content
   • Supports markdown formatting
   • Right-click to toggle between formatted and plain text views

Status Bar

• Shows current operation status
• Displays error messages
• Indicates clipboard updates
• Shows LLM operation progress

Configuration

config.json

{
    "OLLAMA_URL": "http://localhost:11434/api/",
    "DEFAULT_MODEL": "aya-expanse:latest"
}

prompts.json

Contains transformation templates for different operations. Each template can include:

• Description
• System prompt
• User prompt template
• Example inputs/outputs

Adding New Prompts

The prompts.json is used to store the different prompts that can be applied to the clipboard content. Each key in the dictionary represents the prompt name, and the corresponding value is the prompt template.

Step-by-Step Guide to Add New Prompts

1. Open the prompts.json file in your preferred code editor.
2. Add a new key-value pair to the dictionary, where the key is the name of the prompt, and the value is the prompt template.
3. Run python run.py

Example

To add a new prompt for rewriting the text in short sentences, you can modify the prompts.json file as follows:

{
    "Chat Mode": "\"{}\"",
    "Rephrase": "Please rephrase the following text while keeping the original meaning without any preamble: \"{}\"",
    "Translate in English": "Please translate the following text into English without any preamble: \"{}\"",
    "Summarize": "Please summarize the following text: \"{}\"",
    "Rephrase in short sentences": "Please rephrase the following text in short sentences while keeping the original meaning without any preamble: \"{}\""
}

Building Executables

To build standalone executables for Windows, Linux, or macOS:

1. Install the required dependencies:

   pip install -r requirements.txt

2. Run the build script:

   python build.py

The executables will be created in the dist directory, organized by platform:

• Windows: dist/windows/ClipAI.exe
• Linux: dist/linux/ClipAI
• macOS: dist/darwin/ClipAI

Each platform directory contains:

• The executable file
• Required configuration files (config.json, prompts.json)
• Images directory with all icons
• README.md

Note: To build for a specific platform, you need to run the build script on that platform. Cross-platform building is not supported.

Project Structure

run.py
build.py
config.json
prompts.json
src/
├── __init__.py
├── main.py
├── core/
│   ├── __init__.py
│   ├── config.py
│   ├── error_handler.py
│   ├── llm_client.py
│   └── markdown_parser.py
└── ui/
    ├── __init__.py
    ├── clipboard_viewer.py
    └── components.py

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.