AnyModelBot on Railway

Open-source personal AI bots on any model, with memory and chat channels, on Railway.

Official site · License: MIT · 1 Railway template

Deploy AnyModelBot on RailwayCompare templates

AnyModelBot is an open-source app for personal AI bots. Each bot has its own persona, model, workspace and memory, works on tasks in the background, and answers you in a browser console or in Telegram, Slack, Discord, Feishu or WeChat.

It is local-first: a Node.js server with a web console, SQLite, a bundled model gateway and an agent runtime that can run commands and edit files inside each bot's workspace. Running it on a server means scheduled rules, background tasks and chat channels keep working when your own computer is off.

Bots use either your own API key from a model provider or a Modelflare account, a pay-per-use gateway from the same GitHub organization. AnyModelBot is released under the MIT licence.

What AnyModelBot can do

  • Separate bots with their own persona, model, workspace and memory
  • Web console plus Telegram, Slack, Discord, Feishu and WeChat channels
  • Scheduled rules and background tasks
  • Agent runtime that runs commands and edits files in a bot workspace
  • Bring your own provider API key or use a Modelflare account
  • Group chats with several bots on different models

What people use it for

  • A personal assistant you message from Telegram, Slack or Discord
  • Scheduled checks that only notify you when something changes
  • Background research and reports delivered to your chat
  • Running several bots on different models side by side

Videos

AnyModelBot GitHub Explained: Local Memory, Sandboxed Workspaces, and Model Routing · Alex Hitt

Running AnyModelBot on Railway

The Railway template runs AnyModelBot as a single service from the upstream image with a small start wrapper, plus a volume at /data that holds chats, memory, workspaces, channel tokens and model keys. The console only answers requests that carry a generated access token. There is no separate database service. Railway bills the CPU, memory and volume the service uses; model usage is billed by the provider or by Modelflare, not by Railway.

AnyModelBot templates on Railway

AnyModelBot Personal AI Bots on Any ModelMaintained by Tracklayer
AnyModelBot is an MIT-licensed app for personal AI bots, each with its own persona, model, workspace and memory. This template runs v0.1.23 as an always-on server behind a generate
1 deployments
Deploy

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

Deployment numbers as of 2026-11-20.

Questions about AnyModelBot

What do I need before deploying?

A model source: either your own API key from a supported provider such as OpenAI, Anthropic or Google Gemini, or a Modelflare account. Telegram, Slack, Discord, Feishu or WeChat access is optional and set up afterwards.

How do I sign in to the console?

The console only answers requests that carry an access token that the template generates at deploy. Find it in the service variables and keep it private.

Where is my data stored?

On the Railway volume at /data: chats, memory, workspaces, channel tokens and model keys. Back up that volume if you rely on it.

Who bills for model usage?

With your own API key, requests go straight to that provider, which bills you. With a Modelflare account, Modelflare bills you; image and video generation and per-bot budgets are only available that way.

Similar apps

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