The Jac Programming Language and Jaseci Stack

SYS://JASECI

One ecosystem forAI-native software.

Jaseci is the home of Jac: a full-stack programming language with AI built in, a hackathon series where its community ships, and a home where your creations are hosted and scaled.

01 / jaclang.org

One language.
One compiler.The whole stack.

Jac builds AI, graphs, persistence, APIs, UIs, and cloud deployment directly into the language, as a native superset of Python with the PyPI and npm ecosystems at your fingertips.

AI as a language feature

Hand any function to an LLM with a single keyword: by llm. No prompt plumbing, no SDK ceremony.

Object-Spatial Programming

Nodes, edges, and walkers make graphs and the computation that moves over them first-class citizens.

The whole stack, one program

Backend, frontend, persistence, and APIs come out of one codebase. The compiler does the gluing.

Laptop to cloud, zero rewrites

The same program runs as a script, a service, or a Kubernetes deployment. Scale is a flag, not a refactor.

feed.jac
import from mtllm { Model }
glob llm = Model(model_name="gpt-5");

"Rank posts by what this user will love."
def rank(posts: list[Post]) -> list[Post] by llm();

node Post { has text: str; }

walker feed {
    can serve with `root entry {
        report rank([-->(`?Post)]);
    }
}

$ jac start feed.jac # API, persistence, and AI. Served.

SYS://JACHACKSv3.0

$ jachacks --list

EVENT://SFIN-PERSON

San Francisco, CA

A one-day build sprint for the Bay Area's builders.

EVENT://A2IN-PERSON

Ann Arbor, MI

Where the series was born, on campus at Michigan.

EVENT://ONLINEREMOTE

Anywhere on Earth

Global online rounds. All you need is a laptop.

> register at jachacks.org

02 / jachacks.org

Where the Jac
community ships.

JacHacks is the hackathon series powering the future of AI programming. Form a team, learn the stack in a morning, and demo a working AI-native app by night. In person and online, from San Francisco to Ann Arbor to wherever you are.

  • >Build with Jac and the full Jaseci stack
  • >Mentors, workshops, and starter kits included
  • >Prizes, swag, and a fast track into the community

03 / jachammer.ai

A home for yourJac creations.

JacHammer is where your Jac projects live. Host and deploy your creations at scale, share them with a link, and take an idea from prototype to production without ever moving it.

Deploy with a click

Go from working code to a live, hosted app in seconds. Every project gets its own shareable URL.

Scale built in

Hosted on the Jaseci stack, your app grows from weekend demo to production traffic without a rewrite.

Build in your browser

A full web IDE is included. Edit, live-preview, and iterate on your projects without installing a thing.

Every version kept

Project history lives with the project. Roll back, branch, and compare releases without leaving the page.

jachammer.ai/projects

my projects

+ new
tweetspherev23
LIVE
tweetsphere.jachammer.app
open in ide·versions·deploy
study-buddyv8
LIVE
study-buddy.jachammer.app
open in ide·versions·deploy
feed-rankerv2
DEPLOYING
feed-ranker.jachammer.app
open in ide·versions·deploy

all systems live · scaling on demand

community

Hear what builders and thecommunity think about Jac.

Enterprise teams shipping to production, startup founders building on the stack, and first-time Jac developers at JacHacks — in their own words.

Featured

Jaseci is a game changer for ProServe because it allows us to build full stack prototypes quickly and transition them to production with zero friction.

Asim Salim

BCS TechnologiesProServe
Featured

Jaseci's architecture gives our team a framework that feels intentional — it is not glue code on top of an LLM API. The emphasis on simplifying complex AI integration and prompt and model evaluation aligns with how enterprise teams actually think about building maintainable systems.

Pat Rinaldo

Ally BankEnterprise AI practitioner
Featured

I see byLLM as a valuable innovation that could make LLM integration easier to develop, more maintainable and enterprise-ready.

Brian Yang

GenesysEnterprise AI practitioner
Featured

Using Jaseci, our small engineering team has delivered features that would have required significantly more time and infrastructure with conventional tools. The byLLM component has enabled faster iteration, higher consistency, and measurable improvements in AI-driven user experience.

Patrick McLaughlin

Founder of Tobu
JacHacks Participant

I particularly liked the unified full-stack experience, graph-based programming model, and the byLLM keyword for native AI integration. Together, they reduced boilerplate, accelerated development, and made it easier to focus on building intelligent applications rather than managing multiple tools and frameworks.

Sarath Kothapalli

1–2 yrs coding exp.Brand new to Jac
JacHacks Participant

The one language end to end development was great because it made it really easy to build using the language and change ideas with the AI understanding everything it was doing from the start. It was really seamless, the agent understood the language and architecture much easily and wrote better more streamlined code overall.

Aditya Das

5–10 yrs coding exp.Brand new to Jac
JacHacks Participant

The unified programming paradigm and the graph-based model make structuring full-stack AI workflows cleaner and more intuitive.

Rafael Lopez

10+ yrs coding exp.Brand new to Jac
JacHacks Participant

I liked how I could do everything in one language and didn't have to move between frameworks.

Vachan Bhogi

5–10 yrs coding exp.Py · JS/TS · Java · C/C++ · Rust
JacHacks Participant

The graph-based part and the walkers allowed us to come up with a really unique solution for finding multi drug synergies which I found interesting.

Zorawar Brar

3–5 yrs coding exp.Brand new to Jac
JacHacks Participant

Graph-based programming also felt very natural for this problem. Since the system is inherently graph-structured, modeling it directly as nodes and edges made the logic easier to reason about.

Yanshengtian Dong

1–2 yrs coding exp.Brand new to Jac
JacHacks Participant

I just like the different approach of byLLM instead of prompt engineering. It also meant having a more clean and understandable codebase which is a huge plus.

Yousef Al-Wahami

3–5 yrs coding exp.Brand new to Jac
JacHacks Participant

I really like the graph based memory and AI native design. Smooth interaction with other tech stack. Clean agent building.

Nitish Kumar Singh

3–5 yrs coding exp.Built with Jac before10/10 would use again

swipe to explore

Start anywhere. It all runs on Jaseci.

Learn the language, join a hackathon, or give your next creation a home.