Jan

Jan is a versatile and user-centric AI platform designed to be a viable open-source alternative to proprietary services like ChatGPT. Its core philosophy is built on being local-first, user-owned, and highly customizable. This means your conversations and data remain on your device, giving you complete control and privacy. Jan operates as a desktop application for Windows, macOS, and Linux, transforming your personal computer into a powerful, self-contained AI workstation.

With Jan, you can interact with cutting-edge AI without an internet connection. The platform is engineered to run powerful open-source models directly on your hardware. This offline capability is perfect for users concerned about data privacy, those with intermittent internet access, or anyone who prefers to keep their work confidential. The application is built in public, fostering a transparent and community-driven development process.

How to use Jan

Getting started with Jan is straightforward and designed for both beginners and advanced users:

1. Download and Install: Visit the official Jan website (jan.ai) and download the application for your operating system (Windows, macOS, or Linux).
2. Choose a Model: After launching Jan, navigate to the Model Hub. You can download and run popular open-source models like Llama 3, Gemma, or Mistral with a single click. Any GGUF model from Hugging Face is also compatible.
3. Connect to Cloud APIs (Optional): If you need the power of larger, cloud-based models, you can easily connect your API keys from providers like OpenAI, Groq, Anthropic, or Cohere. Jan acts as a unified interface for both local and cloud models.
4. Start Chatting: Once a model is loaded, you can begin chatting, asking questions, brainstorming ideas, or getting assistance with tasks, just like you would with other AI chat tools.
5. Utilize the Local API: For developers, Jan offers a one-click setup for a local, OpenAI-compatible API server running at localhost:1337. This allows you to integrate local LLMs into your own applications and development workflows.
6. Interact with Files (Experimental): Use the experimental feature to chat with your PDFs, notes, and other documents to extract summaries, find answers, or gain insights directly from your files.

Core Features of Jan

• 100% Offline Operation: Run powerful AI models on your device without needing an internet connection, ensuring your data never leaves your computer.
• Integrated Model Hub: Easily discover, download, and manage a wide range of open-source LLMs, including Llama 3, Mistral, Gemma, and Jan's own optimized models.
• Hybrid AI Access: Seamlessly switch between local models for privacy and speed, and powerful cloud-based models (OpenAI, Groq, etc.) for more demanding tasks.
• Local OpenAI-Compatible API Server: A built-in server allows developers to use Jan as a local backend for their AI applications, facilitating rapid prototyping and development without API costs.
• High Customizability: Tailor the application to your needs with extensions, custom model parameters, and adjustable settings for a personalized experience.
• Model Context Protocol (MCP): An advanced feature that enables the AI to interact with external tools and applications, such as browsing the web, analyzing data in Jupyter notebooks, or creating designs in Canva.
• Chat with Documents: An experimental feature that allows you to have conversations with your local files (PDFs, text files) to quickly find information and generate summaries.
• Open Source and Community-Driven: With its code available on GitHub under an Apache 2.0 license, Jan benefits from community contributions and maintains full transparency.

Use Cases for Jan

Jan is suitable for a diverse range of users and applications:

• Software Developers: Use the local API for building and testing AI-powered features, generating code, and debugging without relying on external services.
• Researchers and Academics: Analyze sensitive documents, summarize research papers, and conduct experiments with various LLMs in a secure, private environment.
• Content Creators and Writers: Brainstorm ideas, draft articles, and overcome writer's block, even when offline.
• Privacy-Conscious Individuals: Engage in private conversations with an AI assistant without concerns about data being collected or used by third parties.
• Students: Learn new topics, get help with homework, and explore the capabilities of different AI models for educational purposes.

Advantages of Jan

Jan offers several key advantages over traditional cloud-based AI services:

• Ultimate Privacy and Security: Since everything runs locally, your data is completely private and secure on your own machine.
• No Subscription Fees: The core application and local model usage are entirely free. You only pay for what you use when connecting to third-party cloud APIs.
• Full Data Ownership: Your data is stored in universal formats on your device, with no vendor lock-in. You can move or delete it at any time.
• Uninterrupted Access: Work with AI anytime, anywhere, regardless of internet connectivity.
• Flexibility and Control: Choose from a vast library of open-source models and customize their behavior to fit your specific needs, free from corporate censorship or moderation.

Pricing and Plans

Jan is fundamentally free for its core functionality. The pricing model is simple and transparent:

• Local Use: The Jan desktop application and running local models are completely free and open source. There are no subscriptions, usage limits, or hidden costs.
• Cloud Models: When you connect to cloud AI providers like OpenAI or Anthropic, you pay the provider directly for your API usage. Jan does not add any markup or fees.
• Future Services: Optional paid services, such as a managed Jan cloud platform, are planned for late 2025, but the core local application will always remain free.