Codain CLI

Codain is a simple command-line interface (CLI) that simplifies running and managing local AI models using the Ollama backend. It is lightweight and optimized for environments like Termux/Terminal.

Features

• Easy Setup: A single codain install command sets up ollama, and other core dependencies.
• Smart Process Management: Automatically starts the ollama serve process with codain run and cleans it up with codain stop.
• Model Management: Easily uninstall any downloaded model using codain remove.
• Termux Friendly: Built with a robust installation script (using pkg Or other package also supported like apk, apt, dnf, pacman ) , that runs smoothly in Terminal.

Prerequisites

• Environment: A Linux-based environment.
• Node.js: npm (Node.js) must be installed.
• Hardware: A minimum of 4GB RAM and ~5GB of free storage is recommended.

Installation

1. First, install the package globally via npm:

   npm install -g codain

2. After installation, you must run the one-time setup command:

   codain install

   This command will install ollama and other core dependencies. It will then display an interactive menu for you to download your first AI model.

Usage

Here are the available commands:

1. Run a Model

This command first starts the Ollama server and then runs the specified model.

# Run the default model (qwen2.5-coder:0.5b)
codain run

# Run a specific model
codain r llama3:8b

2. Stop the Server

Stops the ollama serve process running in the background.

codain stop

# Short command (Alias)
codain s

3. Install/Setup Environment

If you skipped the initial setup or need to re-run it, you can use this command.

codain install

# Short command (Alias)
codain i

4. Remove a Model

Uninstalls a downloaded model from Ollama.

codain remove <model_name>

# Example
codain d tinyllama

5. Show Version

Displays the current version of Codain.

codain --version

# Short command (Alias)
codain -v

6. Show Help

Displays the list of all available commands.

codain help

Author

• Dev. Sourav

📄 License

MIT