Llama Workspace
An open-source & extensible ChatGPT Teams alternative
Documentation · Discord · Twitter
About
Llama Workspace is an open-source & extensible ChatGPT Teams/Enterprise alternative. It is designed to help organizations:
- Cut between 50% and 75% the cost of running an AI assistant in the workplace.
- Centralize access to generative AI tools in a single interface.
- Retain full control over sensitive data by hosting the product on their own infrastructure.
Features
- User & permissions management: Manage people access & decide who has access to which apps.
- All Large Language Models in one place: Ask questions to GPT-4, Claude, Mistral or any other major Large Language Model.
- Apps / GPTs: Create apps for repeatable use cases and share them across teams.
- Chat with documents and data: Upload documents and ask questions to them.
- Build AI chats with your own code: Javascript SDK to build your own AI workflows.
Getting started
All the information on how to self host Llama Workspace can be found in our documentation: https://docs.llamaworkspace.ai
If you want to skip the docs and get started right away, here's a tl;dr:
# 1. Clone the repo git clone https://github.com/llamaworkspace/llamaworkspace.git && cd llamaworkspace # 2. Bootstrap the app DEMO_MODE=true docker-compose -f infra/docker-compose.demo.yml up
Tech stack
- NextJS - Framework
- Trpc + Tanstack Query - BFF and type safe client/server communications
- Typescript - Language
- TailwindCSS - Styling
- shadcn/ui - UI components
- Prisma - ORM
- PostgreSQL - Database
- NextAuth.js - Authentication
- BullMQ + LlamaQ (custom-made service) + Redis - Workers and queues management
- Minio - Object storage
Feedback
We are happy to hear your valuable feedback. Reach out to us in the official Discord channel or on X (formerly Twitter).
License & Trademarks
Llama Workspace is open source under the GNU Affero General Public License Version 3 (AGPLv3) or any later version.