OpenClaw
Open-source personal AI assistant on your own machine: answers in Telegram, Slack, Discord and WhatsApp, extended with skills and plugins
Install
curl -fsSL https://openclaw.ai/install.sh | bashmacOS, Linux and WSL2. The installer starts onboarding. For Windows PowerShell: iwr -useb https://openclaw.ai/install.ps1 | iex.
This is third-party code. Review the repository files before installing.
What it does
OpenClaw runs a local gateway on your computer or server that holds sessions, tools and messenger connections. The assistant is reachable in Telegram, WhatsApp, Slack, Discord, Signal, iMessage and other channels, plus a web control panel, a TUI and apps for macOS, iOS, Android, Windows and Linux. Models and agent harnesses such as Claude and Codex plug in as swappable providers. Capabilities grow through skills, tools and plugins from the ClawHub catalog, and the repository itself ships extensions for the browser, Slack, Feishu and other services.
Who it is for. For people who want to run their own assistant on their own hardware and talk to it through familiar messengers, solo or as a team.
Good fit when
- You want an assistant that answers in Telegram or Slack and works with your files and services
- You want memory, state and credentials to stay on your hardware
- You need a shared team deployment
Not a fit when
- You only need a coding assistant inside an editor
- You cannot keep a gateway running
Example request
Every morning send me a Telegram summary of unread email and today's tasksLimitations
Requires Node.js 24.16+ or 26.1+ (the installer provisions it) and a key for at least one model provider. Availability from Russia depends on the chosen provider and messengers. The assistant gets access to channels and tools, so review permissions per the Security docs.
How to disable. Stop the gateway and remove the package with npm uninstall -g openclaw. Full service removal is described at docs.openclaw.ai.
Security check
- The assistant sends messages in messengers on your behalf
- Runs tools and commands on the gateway host
- Stores provider keys and channel tokens
README in short
The README covers an install script for macOS, Linux and Windows or npm, followed by the onboard wizard that checks model access and configures the gateway. The architecture: the gateway as control plane, web panel, CLI and TUI as clients, channels for messengers, and companion apps with voice and camera. Telemetry and the security model are explained separately. MIT licensed.
FAQ
Is there a paid tier?
Per the README, no: the nonprofit OpenClaw Foundation runs it with no hosted or paid version. You only pay your model provider.
What leaves the machine?
Prompts go to your model provider and messengers. OpenClaw itself only checks for updates by default; statistics are opt-in.
Related
A self-improving agent from Nous Research with a TUI, messaging gateway, cron jobs and skills it writes itself
Anthropic's coding agent for the terminal, IDE and GitHub: understands your codebase, runs tasks and handles git
OpenAI's coding agent that runs locally in your terminal, with IDE and desktop versions
Google's open-source terminal agent on Gemini models: code work, shell, web search and MCP