⚛️ Axion CLI

The Free, Open-Source AI Coding Assistant for your Terminal

Stop paying for AI coding assistants. Run them for free, locally, on your own hardware.

Installation • Quick Start • Features • Configuration • Contributing

---

⚛️ What is Axion?

Axion is a powerful, terminal-based AI coding assistant designed to give individuals and teams the power of Claude Code without the subscription fees. It connects natively to local LLMs (via Ollama) or free-tier APIs (like Google Gemini or Groq) to help you write, edit, and understand code in real-time.

✨ Features

Feature	Description
🤖 Local-First	Pull and run any model from Ollama (Llama 3, Qwen, DeepSeek) for 100% free, private usage.
🛠️ Tool-Equipped	Can read, write, and surgically edit files (edit_file with search-and-replace).
🐚 Terminal Power	Execute shell commands directly and capture their output for debugging.
🎨 Rich UI	Beautiful markdown rendering, syntax highlighting, and progress indicators.
🔒 Safety Built-in	Every destructive operation requires explicit user permission.
⚡ Blazing Fast	Optimized for zero-latency streaming on your local GPU.

🛠️ Tools at your Command

Axion isn't just a chat bot; it's an agent that can:

• 📖 Read File: Analyzes existing code to provide context-aware help.
• ✍️ Write File: Generates entire new modules or files from scratch.
• ✂️ Edit File: Performs surgical edits using a specialized search-and-replace algorithm to minimize mistakes.
• 💻 Run Command: Compiles code, runs tests, or manages your environment.
• 🔍 Search Code: Uses lightning-fast search to find exactly what you're looking for.
• 📁 List Directory: Explores your codebase structure automatically.

🚀 Installation

1. Requirements

• Python 3.9 or higher
• Ollama installed and running (for local usage)

2. Setup

# Clone the repository
git clone https://github.com/iam-ssrivastav/axion-cli.git
cd axion-cli

# Install locally
pip install -e .

⌨️ Quick Start

Start Chatting

Just type axion anywhere in your terminal to start a session in your current directory.

axion

Configure your Engine

Don't worry about editing YAML files. Use the built-in config commands:

# Set your engine to Ollama
axion config set provider ollama

# Set your model (e.g., Llama 3.1)
axion config set model llama3.1:8b

# Using Gemini? Set your key
export GEMINI_API_KEY="your-key-here"
axion config set provider gemini

📜 Commands

Command	Action
/help	List all available slash commands
/config	View or change your AI settings
/clear	Start a fresh conversation context
/model <name>	Hot-swap the model mid-conversation
/exit	Safely end the session

🤝 Contributing

We love forks! 🍴 Whether it's adding a new provider like Anthropic or building a tool to deploy to AWS, we welcome all PRs.

Check out our Contributing Guide to get started.

⭐ Support the Project

If Axion helps you code faster and save money, please star the repo and share it with your friends!

---

Built with ❤️ for the Open Source Community.