Christopher is a local AI chatbot. Get started in a few simple steps.

bash ./setup.sh

chmod +x ./setup-ui.sh
./setup-ui.sh

./setup-ui.sh <host-ip>

.\setup.ps1

.\setup-ui.ps1

At the heart of Christopher lies Ollama, a revolutionary tool that brings large language models (LLMs) to your local machine.

Running on Low-End Hardware

When you run the setup scripts (setup.sh, setup.ps1, or visual loader variants), Christopher auto-pulls the default model (llama3.2:1b) if it is missing. These models are *quantized to 4-bit to fit into limited RAM (as low as 4-8 GB).

Despite their smaller size, these models are still powerful enough for general conversation, coding help, and creative tasks. The quantization process allows them to run efficiently on a laptop with no GPU and as little as 4-8 GB of RAM, without sacrificing too much accuracy.

Christopher lives inside a Docker container 🐳. This isolates the AI environment from your host OS, ensuring stability.

Why Docker🐳?

• 🛅Dependency Isolation: Christopher brings its own Python, Node.js, and system libraries.
• 🗝️Security: Docker provides an additional layer of security by isolating the AI environment from the host system.
• 🤓Ease of Use: 🐳 Docker simplifies the deployment and management of Christopher, making it accessible to users without deep technical knowledge.
• 🆕Reproducibility: The source distribution of Christopher from this site or GitHub runs bit-for-bit identically on any other machine.
• 🗑️Clean Uninstall: Want to remove Christopher? Just remove the container.
Advanced Features:
• ⚙️Advanced User Control: Docker allows you to change GPU, CPU, and RAM usage as needed. Change these settings based on your hardware capabilities.
• 🚀Portability: Move the container to another machine without complex setup.
• ↗️Ports: Docker allows you to map ports for external access.
*Note: Exposing ports can introduce security risks. Always ensure proper firewall settings and access controls are in place.
• ♾️Connectivity: Docker enables seamless integration with other services and tools. The creator of Christopher is passionate about open source🔓, the Docker folder and files are yours to modify.♥️
*Note: Docker is a powerful tool, but it may have a learning curve for those new to containerization. The creator of Christopher is available to help with any questions or issues you may encounter.

Christopher is a Local Area Network (LAN) service. It runs on your own hardware and is meant to be opened directly from the Host or other devices on the same network.

After setup, the default way to connect is the secure LAN URL printed by the installer, such as https://<host-ip>:3001. No DNS records or host file edits are required for the basic deployment.

That design keeps the project approachable for people who just want to install, open one address, and start chatting without any extra network plumbing.

The face of Christopher is a modern, responsive web application built using Next.js. It is optimized for quick load times, clean interaction, and a simple chat-first workflow.

Key Technologies

• React: Updates the chat window instantly without reloading the page and keeps the interface responsive while messages stream.
• Next.js: Handles the browser experience, routes, and local API integration so the app feels like a single coherent product.
• Tailwind CSS: Used for the cyberpunk styling, spacing, and responsive layout behavior across screens.

If you want to contribute to the user experience, we especially welcome ideas that make the interface feel more native, faster, or easier to use on touch devices.

1. PREAMBLE

These Terms and Conditions ("Terms") govern your use of the Christopher AI software ("the Software"), a self-hosted, local-area-network (LAN) artificial intelligence chatbot powered by Ollama and Next.js. By downloading, installing, or using the Software, you agree to be bound by these Terms.

2. LICENSE GRANT (OPEN SOURCE)

Subject to your compliance with these Terms, the Project Creator grants you a perpetual, worldwide, non-exclusive, royalty-free, irrevocable license under the Project Creator's copyright to:

1. Use: Run the Software for personal, educational, or commercial purposes.
2. Modify: Alter the source code to suit your needs.
3. Distribute: Copy and distribute the Software or modified versions.
4. Sublicense: Grant sublicenses to others.

Base License: This Software is licensed under the GNU Affero General Public License v3.0 (AGPL v3). This license ensures that any modifications or derivative works must also be released under the same license, promoting full transparency and community sharing.

View the full AGPL v3 license text

3. DEFINITIONS

• "Host Device": The physical machine (server, laptop, desktop) where the Docker container is running.
• "User": Any individual accessing the Christopher AI interface via a web browser on the local network.
• "AI Output": Any text, code, image, or data generated by the underlying Large Language Model (LLM) (e.g., Llama 3) running within the container.
• "Local Storage": Data stored in the browser's localStorage or Docker volumes on the Host Device.
• "Project Creator": The original author(s) and contributors of the Christopher AI source code.

4. NATURE OF THE SERVICE & LOCAL DEPLOYMENT

5. DATA PRIVACY & IMPERMANENCE

• No Data Collection: The Project Creator does not collect, track, or store any user data, chat logs, or usage statistics. The Software contains no telemetry.
• Network Security: Since the Software runs on your local network, you are responsible for securing your network against unauthorized access. The Project Creator accepts no liability for data exposure resulting from compromised local networks.

6. ARTIFICIAL INTELLIGENCE OUTPUT DISCLAIMER

• Hallucinations: AI models may generate factually incorrect, misleading, or nonsensical information ("hallucinations"). You acknowledge that the Software may produce errors and agree to verify all critical information independently.
• Content Moderation: Christopher may include baseline safety instructions at the application layer, but output boundaries are still heavily influenced by the selected model. Depending on model configuration and training, the AI may still generate offensive, biased, or inappropriate content. You remain responsible for configuring deployment behavior to align with local laws and your ethical standards.

7. INTELLECTUAL PROPERTY & MODELS

• Software Code: The source code of Christopher AI is open source under AGPL v3.
• Third-Party Models: The underlying AI models (e.g., Llama 3, Mistral) used by the Software are subject to their own separate licenses (e.g., Meta's Llama Community License). You must comply with the terms of the specific model you choose to run.
• Trademark: "Christopher AI" is used descriptively. No trademark claims are made. You may fork and rename the project, but should not use the name to confuse users about the origin of derivative works.

8. LIMITATION OF LIABILITY

9. INDEMNIFICATION

You agree to indemnify, defend, and hold harmless the Project Creator from any claims, liabilities, damages, losses, or expenses (including legal fees) arising out of your use of the Software, your violation of these Terms, or your deployment of the Software in a manner that violates applicable laws.

10. TERMINATION

These Terms shall remain in effect until terminated. If you breach any provision of these Terms, your license to use the Software terminates automatically. Upon termination, you must cease all use of the Software and destroy all copies of the source code in your possession.

11. GOVERNING LAW

These Terms shall be governed by and construed in accordance with the laws of Great Britain, without regard to its conflict of law provisions.

Christopher is a self-hosted, multi-user cyberpunk AI chatbot powered by Ollama and Next.js. It brings the power of Large Language Models (LLMs) to your local network without sending a single byte of data to the cloud. Designed for privacy enthusiasts, developers, and home-labbers who demand sovereignty over their digital conversations.

🖥️ System Requirements

Christopher is optimized to run on consumer hardware, but performance scales with your specs.

🚀 Installation

Prerequisites: Installing Docker

Before running Christopher, you must have Docker installed on your system. Choose the guide below for your operating system:

• Windows: Download and install Docker Desktop for Windows. Ensure "WSL 2 backend" is enabled during installation for best performance.
• macOS: Download and install Docker Desktop for Mac. Select the appropriate version for your chip (Intel or Apple Silicon).
• Linux (Ubuntu/Debian): Open a terminal and run the official installation script:

  curl -fsSL https://get.docker.com -o get-docker.sh && sudo sh get-docker.sh

  Add your user to the docker group: sudo usermod -aG docker $USER (then log out and back in).

1. Clone & Start

Open a terminal in the project folder and run the matching setup command for your operating system:

• Linux/macOS: bash ./setup.sh
• Linux/macOS visual loader: chmod +x ./setup-ui.sh then ./setup-ui.sh
• macOS manual fallback: ./setup-ui.sh <host-ip> if interface detection misses your adapter
• Windows: .\setup.ps1
• Windows visual loader: .\setup-ui.ps1

2. Wait for Initialization

The first run may take several minutes while the AI model (~1.2 GB) downloads.

Watch progress: docker compose -p christopher logs -f ollama

Confirm model availability: docker compose -p christopher exec -T ollama ollama list

Ensure llama3.2:1b appears before first chat.

3. Launch the App

This project uses a simple reverse proxy so the app works immediately over your LAN IP.

1. Run the matching setup command for your OS from the project root.
2. No DNS setup is required for the default install.
3. Open the secure LAN URL printed by the setup script, such as https://<host-ip>:3001.

4. Launch Interface

Open the secure LAN IP URL in your browser.

Example: https://<host-ip>:3001

5. Cross-Device Use

• Works across Windows, macOS, Linux, Android, and iOS clients on the same LAN.
• Only the host machine needs Docker; client devices only need a browser.
• Each client may need to trust the generated certificate on first use.

🔐 First Time Setup

User Accounts

Login: Create/select a profile with a required passphrase.
Example: Neo / CorrectHorseBatteryStaple

Profiles: You can create multiple profiles (e.g., "Work", "Personal", "Dev") to keep chat histories completely separate.

Performance Expectations

• Cold Start: The very first message may take 10-20 seconds to generate while the model loads into RAM.
• Streaming: Subsequent messages will stream instantly (tokens per second depends on CPU/GPU).

⚠️ Data Privacy & Impermanence

✨ Features

• 🔒 100% Offline: No API keys, no cloud subscriptions, no telemetry.
• 👥 Multi-User: Different usernames create completely separate chat histories.
• 🛡️ Security-Forward Profiles: Password-protected profiles keep local chat storage encrypted at rest.
• 📶 Transport Awareness Badge: Runtime header indicator shows secure HTTPS mode vs HTTP fallback mode.
• 🧵 Multi-Thread: Click the + in the sidebar to start new topics (e.g., "Coding", "Creative Writing", "Recipes").
• ✏️ Rename Threads: Double-click any thread name in the sidebar to rename it.
• 🎨 Cyberpunk UI: Fully responsive, dark-mode, neon-themed interface optimized for low-light environments.
• 📱 Mobile Ready: Access from any device on your local network (phone, tablet, laptop) using the host IP.

🛑 Maintenance

Stop the Service

To stop the AI and free up resources:

docker compose down

(Your downloaded models and chat history are saved in Docker volumes and will persist when you start it again.)

Update Christopher

docker compose pull
docker compose up -d --build

Remove Completely

docker compose down -v

⚠️ Warning: Removes volumes (deletes all chats & models)

🆘 Troubleshooting

🏗️ Architecture

• Frontend: Next.js 14, React, Tailwind CSS
• AI Engine: Ollama (serving Llama 3 1B/3B Quantized)
• Deployment: Docker Compose + Caddy (LAN HTTPS reverse proxy)
• Storage: Browser localStorage (Client-side) + Docker Volumes (Server-side models)