modelflare-ai/AnyModelBot

What is it?

Key features

• 01Multiple bots, each with its own persona, default model, workspace and memory.
• 02Reach bots from the desktop app, command line, Docker, WeChat, Feishu, Telegram, Slack and Discord.
• 03Access GPT, Claude, Gemini, Grok, DeepSeek and 70+ other models through one Modelflare key.
• 04Three permission levels per bot, with approval cards for actions outside the workspace.
• 05Background tasks and scheduled rules with per-task budgets and quiet hours.
• 06Built-in computer use via Cua Driver, with approval required for each click and keystroke when enabled.

Use cases

• Running a personal AI assistant on your own computer that answers in WeChat or Feishu.
• Letting a bot organize files in a folder while you approve any moves outside it.
• Scheduling periodic checks, such as monitoring a price and notifying you when it crosses a threshold.
• Comparing and switching between different LLM providers without maintaining separate subscriptions.

Quick start

1. 1

   Download the installer for your platform from the Releases page

2. 2

   Run the project from source: clone the repository

   git clone https://github.com/modelflare-ai/AnyModelBot.git
3. 3

   Install dependencies and build the bundled gateway

   npm install && npm run setup
4. 4

   Start the local admin UI at http://127.0.0.1:3600/

   npm start

Things to know

• !Using models requires a Modelflare account, and model requests are forwarded through Modelflare to the chosen provider.
• !Building from source needs Node.js 22.13+ and Go 1.26.8+.
• !Computer-use features are Beta on macOS and require a desktop session.

In collections

Topics

FAQ

What is AnyModelBot?

A local-first desktop and Docker AI agent app where you run personal bots that use any model via one Modelflare key, reachable from WeChat, Feishu, Telegram, Slack and Discord. AnyModelBot is a local-first AI agent application for Windows, macOS and Linux, also deployable via Docker. Each bot has its own persona, default model, workspace and memory. You can chat with bots inside the app or reach them from WeChat, Feishu, Telegram, Slack and Discord. Bots can read and write files, run commands in a sandbox, and, with the built-in Cua Driver, operate the desktop; actions that leave the workspace require approval.

Is AnyModelBot free to use? What license does it have?

AnyModelBot is open source under the MIT license. Check the license terms before using it in commercial products.

Is AnyModelBot still maintained?

Actively maintained. The last commit to AnyModelBot was on Nov 6, 2026.

What can I use AnyModelBot for?

Running a personal AI assistant on your own computer that answers in WeChat or Feishu. Letting a bot organize files in a folder while you approve any moves outside it. Scheduling periodic checks, such as monitoring a price and notifying you when it crosses a threshold. Comparing and switching between different LLM providers without maintaining separate subscriptions.

How do I install AnyModelBot?

Download the installer for your platform from the Releases page Run the project from source: clone the repository: git clone https://github.com/modelflare-ai/AnyModelBot.git Install dependencies and build the bundled gateway: npm install && npm run setup Start the local admin UI at http://127.0.0.1:3600/: npm start

Where can I get AnyModelBot?

The source code and installation instructions are on GitHub: https://github.com/modelflare-ai/AnyModelBot