Skip to content

Welcome to Jac#

One Language for AI-Native Full-Stack Development

Jac is a programming language and superset of both Python and TypeScript/JavaScript, with novel constructs for AI-integrated programming. Access the entire PyPI and npm ecosystems while using features like by llm() to seamlessly weave AI into your code. Write backend logic, frontend interfaces, and AI integrations in a single unified language.

Why Jac?#

What You Want How Jac Helps
Build AI apps easily Native LLM integration with by llm() - no prompt engineering
Full-stack in one language React-style frontend + Python backend, seamlessly connected
Use existing libraries Full access to PyPI and npm ecosystems
Deploy without DevOps jac start --scale deploys to Kubernetes automatically
Model complex domains Graph-based Object-Spatial Programming for connected data
Code with AI assistance Clean syntax designed for both humans and AI models to read and write

See it in action

Want to see exactly how much code Jac eliminates? Check out Jac vs Traditional Stack - a side-by-side comparison showing ~30 lines of Jac vs >300 lines of Python + FastAPI + SQLite + TypeScript + React for the same Todo app.

Get Started in 5 Minutes#

Step 1: Install#

pip install jaseci

This installs the complete Jac ecosystem: jaclang (compiler), byllm (AI integration), jac-client (frontend), jac-scale (deployment), and jac-super (enhanced console).

Verify your installation:

jac --version

This also warms the cache, making subsequent commands faster.

Step 2: Create Your First Program#

Create hello.jac:

with entry {
    print("Hello from Jac!");
}

Step 3: Run It#

jac hello.jac

Note: jac is shorthand for jac run - both work identically.

That's it! You just ran your first Jac program.

Choose Your Path#

Core Principles#

Jac is built on six key principles:

  1. AI-Native - LLMs as first-class types via Meaning Typed Programming. Call AI like a function.

  2. Full-Stack in One Language - Write React components alongside server code. No context switching.

  3. Python & JavaScript Superset - Use numpy, pandas, react, tailwind directly. Your existing knowledge applies.

  4. Object-Spatial Programming - Model domains as graphs. Deploy walkers to traverse and transform data.

  5. Cloud-Native - One command to production: jac start --scale handles Kubernetes, Redis, MongoDB.

  6. Human + AI Readable - Clean syntax that both developers and AI models can read and write effectively.

Who is Jac For?#

Jac is designed for developers who want to build AI-powered applications without the complexity of managing multiple languages and tools.

You Are Jac Gives You
Startup Founder Build and ship complete products faster with one language
AI/ML Engineer Native LLM integration without prompt engineering overhead
Full-Stack Developer React frontend + Python backend, no context switching
Python Developer Familiar syntax with powerful new capabilities
Frontend Engineer Write UI components with full access to npm ecosystem
Student/Learner Modern language designed for clarity and simplicity

When to Use Jac#

Jac excels at:

  • AI-powered applications with LLM integration
  • Full-stack web applications (frontend + backend)
  • Applications with complex relational data (graphs, networks)
  • Rapid prototyping with production scalability
  • Projects requiring both Python and JavaScript ecosystems

Consider alternatives for:

  • Performance-critical systems programming (use Rust, C++)
  • Mobile native apps (use Swift, Kotlin)
  • Simple scripts where Python suffices
Resource Description
Installation Detailed setup with IDE configuration
Hello World Your first Jac program (2 min)
First Full-Stack AI App Complete tutorial: API, frontend, AI, walkers (15 min)
Tutorials In-depth learning paths
Language Reference Complete language documentation
CLI Reference All jac commands

Need Help?#