Clawbot
Not Just Another Chatbot. A Personal AI Infrastructure.
Most AI lives in someone else's cloud. clawbot is different; it runs on yours, autonomously, and private by default.clawbot.ai: personal AI, no compromise.
The Evolution From Passive to Proactive AI
Understanding why clawbot represents a fundamental shift in how we interact with artificial intelligence
Generation 1: The Q&A Era
Traditional AI assistants like ChatGPT and Siri revolutionized how we access information, but they remained fundamentally reactive. You ask, they answer. They can't initiate contact, monitor your systems, or execute tasks while you're away. Every interaction requires your explicit request.
Generation 2: The Cloud Dependency Trap
As AI capabilities expanded, so did their appetite for your data. Every conversation, every document, every workflow flows through corporate servers. You gained convenience but surrendered control. Your AI assistant knows everything about you, and so does the company hosting it.
Generation 3: The Autonomous Revolution
clawbot represents the next generation: AI that operates as your personal infrastructure. It runs on hardware you control, acts autonomously on your behalf, integrates deeply with your digital environment, and maintains conversation history across 15+ messaging platforms. This isn't an assistant that waits, it's one that anticipates, monitors, and executes.
Capabilities That Redefine Personal AI
Built on principles of autonomy, privacy, and genuine utility, not marketing buzzwords
Data Sovereignty by Design
Every conversation, file, and workflow stays on infrastructure you own. No corporate servers mining your interactions. No surprise terms-of-service changes. Your AI, your rules, your hardware.
Genuine Task Execution
Unlike chatbots that merely suggest actions, clawbot directly manipulates your system: executing shell commands, writing code, controlling browsers, accessing cameras, and utilizing location services. If Node.js can do it, clawbot can orchestrate it.
Unified Multi-Platform Presence
Communicate with your AI through WhatsApp during commutes, Telegram at your desk, Discord with your team, or Slack at work, with seamless conversation continuity. Your AI lives where you already communicate.
Extensible Skills Ecosystem
The ClawdHub marketplace hosts 565+ community-developed skills following the AgentSkills open standard. Install pre-built workflows or develop custom integrations that precisely match your unique requirements.
Model-Agnostic Architecture
Choose your intelligence backbone: Anthropic's Claude for nuanced reasoning, OpenAI's GPT for versatility, or completely free local models via Ollama. Switch providers without reconfiguring your entire setup.
Defense-Grade Security
Sandboxed execution environments, granular permission controls, and user-defined trust boundaries ensure automation runs safely. 84,000+ developers have audited the open-source codebase. Transparency is our security model.
Real Automation for Real Work
What becomes possible when your AI can genuinely act instead of merely advise
Intelligent Inbox
Automatically categorize emails, draft context-aware responses, and purge promotional clutter
Temporal Orchestration
Schedule meetings by parsing natural language, set conditional reminders, coordinate across timezones
Smart Environment
Integrate with Home Assistant to control IoT devices through conversational commands
DevOps Automation
Review pull requests, manage deployments, monitor system health, troubleshoot production issues
Data Analysis
Process spreadsheets, generate visual reports, extract insights from unstructured data sources
Travel Management
Automated flight check-ins, itinerary organization, real-time travel alert monitoring
Architecture for Developers Who Care
A technical foundation designed for extensibility, transparency, and developer autonomy
Core Architecture
- Gateway-centric control plane
- Distributed node system for multi-device access
- WebSocket-based real-time communication
- Markdown-backed persistent memory
Technology Foundation
- Node.js 22+ runtime environment
- TypeScript for type-safe development
- Docker containerization support
- Tailscale for secure remote access
AI Model Compatibility
- Anthropic Claude (all variants)
- OpenAI GPT-4 and successors
- Ollama for local model inference
- LM Studio integration layer
Platform Integration
- WhatsApp via Baileys protocol
- Telegram through grammY framework
- Discord, Slack, iMessage native support
- Signal, Matrix, Mattermost, 15+ total
How clawbot Orchestrates Intelligence
A simplified view of how messages flow through your personal AI infrastructure
Built by Engineers, For Engineers
Open-source software backed by proven technical leadership and an active global community
Peter Steinberger
Founder & Lead Developer
Peter previously founded PSPDFKit, a PDF technology company trusted by Fortune 500 enterprises globally. After successfully exiting, he channeled 15+ years of software engineering experience into building clawbot, a vision for AI that prioritizes user autonomy, privacy sovereignty, and genuine capability over marketing hyperbole.
- ✓ Architected enterprise-grade developer tools serving millions of users
- ✓ Deep expertise in cross-platform development (macOS, iOS, Linux, Web)
- ✓ Committed to open-source principles and community-driven development
- ✓ Advocates for user privacy and data sovereignty in the AI era
Get Started in Minutes
Download the latest stable release for macOS or deploy via Docker on any platform
Linux and Windows users: Deploy via Docker or WSL2. See installation guide →
Questions From the Community
Answers to common concerns about privacy, capability, and getting started
Unlike cloud-based AI that only responds to questions, clawbot runs on your own infrastructure and can actively execute tasks on your behalf, managing files, controlling applications, integrating with your services, and operating autonomously in the background. You own your data completely.
clawbot itself is 100% free and open-source under MIT License. You only pay for the AI model API you choose (Claude, GPT-4) or run completely free local models via Ollama. There are no subscriptions, no feature gates, no vendor lock-in.
clawbot executes entirely on hardware you control. Your conversations, documents, and data never touch third-party servers unless you explicitly configure external AI APIs. Unlike cloud AI services, there's no company mining your interactions for training data.
clawbot can execute shell commands, read and write files, control your web browser, monitor system resources, integrate with 50+ services, send you proactive notifications, and run scheduled tasks, all while you're offline. It's an AI that genuinely acts, not just advises.
clawbot runs natively on macOS and Linux. Windows users can deploy via Docker or WSL2. You can also host it on any cloud VM (AWS, DigitalOcean, Hetzner) for remote access from any device.
Yes! clawbot integrates seamlessly with WhatsApp, Telegram, Discord, Slack, Signal, iMessage, Matrix, and 15+ other platforms. Talk to your AI assistant through apps you already use daily, with unified conversation history across all channels.
For Mac users, it's as simple as downloading a .dmg file and following a guided setup wizard. Linux and Docker deployments require basic command-line familiarity. Most users complete setup in under 30 minutes.
clawbot features sandboxed execution, configurable permission controls, and user-defined trust boundaries. You decide exactly what it can and cannot do. All code is open-source and auditable. 84,000+ developers have reviewed it.
Join the active Discord community or browse comprehensive documentation at docs.clawd.bot. The community is welcoming to both beginners and advanced users.
Join the Personal AI Movement
84,000+ developers are building the future of autonomous, privacy-first artificial intelligence. Be part of it.










