Rakazo on Railway

Self-hosted AI agent for persistent teammates with chat, memory and routines, on Railway.

Official site · License: Apache-2.0 · 7 Railway templates

Deploy Rakazo on RailwayCompare templates

Rakazo is an open-source, self-hosted AI agent for persistent teammates: bots with their own chats, memory, routines and, optionally, their own computers, running on the model you choose.

People use it for a personal assistant that remembers context, scheduled routines such as a morning inbox summary, and bots that browse, research and run commands in an isolated sandbox. Connected apps such as Gmail, Slack and GitHub are available through Composio.

Rakazo is released under the Apache License 2.0.

What Rakazo can do

  • Persistent bots with memory and routines
  • Web, desktop and mobile apps against one server
  • Bot computers through Railway sandboxes, E2B, Daytona or Boat
  • Bring your own model or subscription
  • Separate web, API and worker services
  • Closed sign-up with the owner account pre-created

What people use it for

  • A personal assistant with weekday routines
  • Bots that browse and edit files on their own Linux computer
  • Coding and operations helpers in an isolated sandbox
  • Agents working in connected apps through Composio

Videos

Rakazo GitHub Overview: Self-Hosted AI Agents, Sandboxes, and Security Tradeoffs · Alex Hitt

Running Rakazo on Railway

The Rakazo template runs v0.8.0 as web, API and worker services from the official image with Postgres and a Railway Bucket; only the web service is public. Railway cannot run Docker computers, so bot computers use remote providers such as Railway sandboxes, E2B, Daytona or Boat. The template estimates about 0.9 GB RAM idle (about $10 a month) and $20 to $25 in typical use; computers and models are extra.

Rakazo templates on Railway

Rakazo Persistent AI TeammatesMaintained by Tracklayer
Deploys Rakazo v0.8.0, an open-source self-hosted AI agent for persistent teammates (bots with their own chats, memory, routines and optional computers), as separate web, API and w
10 deployments
Deploy
Rakazo Template by an independent creator
Deploy Rakazo with private computers, PostgreSQL and persistent storage.
63 deployments
Deploy
Rakazo by an independent creator
Persistent AI teammates with memory, routines, and their own computers.
21 deployments
Deploy
Rakazo Suite by an independent creator
Rakazo persistent cloud agent (Grok Bot Alternative). Non forked.
7 deployments
Deploy
Rakazo (Single Container) by an independent creator
Persistent AI teammates with memory and routines. BYO model key
5 deployments
Deploy
Rakazo by an independent creator
Self-hosted Rakazo: AI teammates with memory, routines and a real computer.
1 deployments
Deploy
Rakazo by an independent creator
AI bots with memory and their own computers. Current release, your login.
0 deployments
Deploy

Tracklayer templates are listed first. Other Rakazo templates follow, ordered by total deployments on Railway. How this works.

Deployment numbers as of 2026-11-20.

Questions about Rakazo

How do I connect a model?

After signing in, add a ChatGPT or Claude subscription or an API key under Settings, Models, or set OPENROUTER_API_KEY or ANTHROPIC_API_KEY on the API service.

Does it work without a sandbox provider?

Yes. Rakazo boots and chat works, but bots get no computer until you configure a provider.

Who can sign up?

Public sign-up is closed and the owner account is created before the app goes online.

What do computers cost?

The template's example: a desktop with Chromium at about 2.5 GB and 0.5 vCPU costs roughly $0.21 per hour while running as a Railway sandbox. Other providers bill separately.

Is it affiliated with Rakazo?

No. It is a community template using the official images.

Similar apps

Tracklayer maintains the templates marked as such; Rakazo is a trademark of its owners and this site is not affiliated with them.