rakazo
by elie222
Rakazo is a self-hosted AI agent app that gives you persistent bots with their own memory, routines and conversation history. It runs on the web, desktop and mobile, and lets you use your own AI model and a sandboxed computer for the bots to work in.
- Stars
- 6k
- Licence
- Apache-2.0
- Runs on
- A small server
- Setup
- Moderate
- Language
- TypeScript
- Age
- 76 days
What you can do with it
- Create persistent AI bots that remember past conversations and run routines
- Give bots a sandboxed computer with a browser, terminal and files to complete tasks
- Connect apps and tools through Composio, Pipedream, MCP servers or OpenAPI, and talk to bots by voice
Make money with it
A freelancer could set up and host customised agent bots with connected apps for small businesses as a paid setup and maintenance service.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Run always-on assistant bots that handle routine admin tasks such as research, file work and browsing, which saves staff time
- Keep company data on your own server while choosing cheaper or private AI models instead of paying for a hosted assistant
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
Run the installer script with Docker, the Compose plugin, curl and OpenSSL, then open it in a browser and connect a model; a Railway template and desktop and mobile apps are also available.
Follow the official instructions in the project README.
Things to check
- Beta software, so expect rough edges
- You need your own model API keys, which cost money, and some remote sandbox providers and voice services need paid keys
- Gives bots terminal and browser access, so sandboxing and server hardening matter
- Treg integration is usage-metered and has terms that restrict hosted resale
Checked 27 Oct 2026. Summary drafted by AI from the project's README.