How to Install and Use the Jan Open Source AI Assistant

Thursday, Jan 2, 2025 | 6 minute read

GitHub Trend

A groundbreaking local AI assistant prioritizing user privacy and control. 🌍 With offline capabilities, rich customization, and community involvement, it empowers users while ensuring data security. 🚀 Enjoy seamless integration from popular models for a tailored experience! 🎉

In this digital age, privacy and security have become top concerns for many users. 🎭 In this context, Jan emerges as a revolutionary local AI assistant. 🌍 Jan is a unique open-source tool dedicated to providing users with unprecedented privacy protection and data security. It places the control of data firmly in the hands of users, allowing them to use it with peace of mind while enjoying the convenience that technology offers. 🎉

What is Jan? — Your Customizable AI Assistant 🎉

Jan is an unprecedented open-source AI assistant designed to offer the highest level of privacy protection and operational control. Unlike many AI assistants that rely on cloud computing, Jan operates entirely locally, disconnected from the internet, ensuring that user data and interactions remain secure and private. Through continuous development and updates, Jan seamlessly delivers the latest features and services to users, significantly enhancing the user experience! ✨

What Makes Jan Different — A Nontraditional AI Experience 🚀

• Offline Capability: Jan runs locally on the user’s device, ensuring data privacy and preventing security risks associated with storing user information on external servers. 🔒
• Model Library: To meet diverse needs, Jan provides several popular language models, including Llama, Gemma, and Mistral, for users to freely choose from. 📚
• Remote AI Connection: While Jan can work independently offline, it also supports connections to cloud AI APIs, such as Groq and OpenRouter, for easy access to additional services and data. 🌐
• Local API Server: Jan can function as a local server similar to OpenAI’s API, supporting local model operations and making it convenient for developers to integrate and invoke AI functionalities. 🛠️
• High Customization: Users can personalize settings through extensions, fully catering to their unique needs and preferences. 🎨

Why Developers Choose Jan? — Putting Technology in User Control 🔑

The appeal of Jan to developers stems from its open-source ecosystem and user sovereignty. With Jan, users not only own their data but can also process it locally, enhancing their ability to exercise control. By participating in the GitHub development community and engaging in real-time interactions on Discord, users can directly influence the future development of the project by submitting suggestions or contributing code for a customized AI experience! 👐

Installing and Using Jan 🚀

To start using Jan, you first need to complete the installation process. Here are some simple steps to help you set up Jan on your device.

Clone the Repository and Set Up 🌐

First, you need to clone Jan’s repository to your local system. Open your terminal and enter the following command:

# Clone the Jan repository
git clone https://github.com/janhq/jan
# Switch to the Jan directory
cd jan
# Check out the desired branch
git checkout -b DESIRED_BRANCH

Note:

• The command git clone https://github.com/janhq/jan downloads the remote Jan project to your local machine, so make sure you have an internet connection! 🌍
• Use cd jan to switch to the newly downloaded Jan folder to prepare for further operations.
• Replace DESIRED_BRANCH in git checkout -b DESIRED_BRANCH with the specific branch name, like main, to ensure you’re accessing the latest features or fixes.

Running the Development Environment and Using the Jan Desktop Application 🖥️

Next, you need to start the development environment. Simply enter the following command:

make dev

Note:

• The command make dev starts the development server, allowing you to access Jan’s desktop application for development and debugging. This process typically takes a few seconds to a few minutes, depending on your system’s performance. ⏳

Building for Production 🏗️

If you want to build the Jan application for production, follow these steps:

# Continue with steps 1 and 2 above
# Build the application
make build

Note:

• Ensure that the previous steps have been successfully completed before running the make build command. This command prepares the application for production, and the final files will be placed in the dist folder. 🌟

Make sure your operating system and hardware meet the requirements to run Jan. Here are some basic system requirements:

• MacOS: Version 13 or above
• Windows:
  • Version 10 or higher
  • Requires Nvidia GPU and CUDA Toolkit 11.7 or higher for GPU support
• Linux:
  • Ensure glibc version is at least 2.27 (verify by ldd --version)
  • gcc, g++, and cpp should be version 11 or higher
  • Also requires installation of Nvidia GPU and CUDA Toolkit for GPU support 💻

Downloading Models 📥

Jan offers several local AI models for you to choose from based on your needs. Here’s how to do it:

1. Visit the Hub.
2. Browse and select the model you need; you can click on the model’s name for more details.
3. If needed, you can enter the Hugging Face model ID or URL in the search box for quick access to the desired model.
4. Select an appropriate model size based on your task requirements to balance performance and resources.
5. Click the Download button to download the model.

Note: These straightforward steps enable you to quickly find and download the required AI models, integrating them into your application. 🚀

Next, you can go to the Thread tab:

1. Click on the Model option.
2. Choose either On-device or Cloud section to decide where your model will run.
3. Adjust the settings as needed to better fit your use case.

Customizing Assistant Commands 🔧

Jan also allows users to customize the assistant’s behavior by specifying queries, commands, or requests in the assistant commands field. This means you can adjust the assistant’s responses for specific purposes, helping to better meet your needs. This feature makes Jan more aligned with your actual requirements, greatly enhancing the user experience! 🎉

By following the above steps, you should be able to successfully install and start using the Jan open-source project, unleashing powerful AI capabilities. 🔑🎈

Getting and Support 📥

• Compatibility: Jan supports various architectures including NVIDIA GPU, Apple M series, Intel Mac, Linux Debian, and Windows x64, ensuring usability across different operating systems. 💻
• Download Options: Users can download the stable, beta, or nightly build versions from the official website to cater to different operating systems such as Windows, MacOS, and Linux. 🌍
• Community Support: Jan offers active development support where users can contribute via GitHub and obtain real-time support and community interaction through Discord. 💬

AI & ML Models — JanHQ 🧠

JanHQ focuses on the development and management of artificial intelligence and machine learning models, providing the latest content and an array of model collections. For different application scenarios, JanHQ offers multiple choices, including the Trinity model for experimental text generation and the Nitro model for advanced text generation tasks, showcasing its robust capabilities and wide-ranging application prospects! 🔍

AI Philosophy and Community Engagement 🌈

Jan emphasizes an open-source ecosystem and is committed to respecting user data privacy through local processing for privacy and security. It encourages users to engage in the development community, submit contributions via GitHub, or join discussions on Discord to suggest new features or report bugs. 🤝

Want to explore and experience the innovative solutions brought by Jan? Visit the official website at jan.ai to download! 🎊