Everything you need to set up and use your KOOMPI Claw. Whether you have the hardware or use the managed service.
Connect your AI assistant to apps, automate workflows, and build custom integrations.
Access multiple LLM providers through a single, unified API. Pay per token, no subscriptions.
Store and retrieve long-term memories with vector search. Your AI remembers context across sessions.
Install and create custom skills. Extend your AI's capabilities with community plugins.
from openclaw import ClawClient
# Initialize client with your API key
client = ClawClient(api_key="your-api-key")
# Send a chat completion request
response = client.chat.complete(
model="glm-5-turbo",
messages=[
{"role": "user", "content": "Hello, KOOMPI Claw!"}
],
memory=True # Enable long-term memory
)
print(response.message)
# Output: Hello! I'm your KOOMPI Claw assistant. How can I help you today?Connect the power adapter and press the power button. OpenClaw starts automatically. Wait ~60 seconds for full initialization.
A QR code will be displayed on the screen. Scan it with your phone camera — this opens the onboarding page.
Register with your email or phone number. This gives you access to the AI Gateway, KConsole, and KStorage.
Add credits to your AI Gateway account. You can start with as little as $1. These credits power your AI conversations.
The onboarding wizard guides you through creating a Telegram bot via @BotFather. Just follow the steps — takes 2 minutes.
Open Telegram and send a message to your bot. Your AI assistant is ready.
Once your bot is working on Telegram, you can easily add WhatsApp, Discord, iMessage, and 16+ other channels through the OpenClaw dashboard. All conversations share the same memory and AI.
Email claw@koompi.cloud or reach out via Telegram. Tell us your requirements — number of users, expected usage, any special needs.
Our team deploys your OpenClaw instance, configures the Telegram bot, sets up memory and monitoring. Usually done within 24 hours.
We send you the Telegram bot link. Start chatting. We handle updates, monitoring, backups, and scaling.
Always running the latest OpenClaw version
Daily backups with 30-day retention
24/7 uptime monitoring and alerts
Optimized for speed. Best for quick questions and casual conversations. Lowest cost per token.
Balanced performance. Good for most tasks. Recommended for general use.
Fast and capable. Supports text and images. Great for visual tasks.
Most capable model. Best for complex reasoning, code, and analysis.
OpenClaw supports 20+ channels including:
You can connect multiple channels simultaneously and your AI will work across all of them with shared memory.
OpenClaw uses a vector database to store long-term memories. It automatically captures important information from conversations (preferences, facts, decisions) and recalls them when relevant.
Your AI remembers your preferences, past conversations, and context. The more you use it, the smarter it gets. All memory is stored locally on your hardware — we never have access to it.
Absolutely! OpenClaw has a skill system (AgentSkills) that lets you install custom capabilities.
The AI needs internet to call the LLM APIs, so it won't respond without a connection.
However, your configuration, memory, and data are safe on your Mini. Once internet returns, everything resumes normally. The system will automatically reconnect and continue working.
Yes! KOOMPI Claw is built on OpenClaw, which is fully open source.
For hardware users, you only pay for AI credits (pay-per-use):
No subscription required. Add credits when you need them. Credits never expire.
Yes! OpenClaw is open source and can run on any Linux server or even a Raspberry Pi.
However, buying a KOOMPI Mini supports development and gives you a plug-and-play experience with guaranteed compatibility.
Complete docs — setup, config, skills, API reference, and more.
Browse and install AgentSkills to extend your AI's capabilities.
Source code, issues, feature requests, and contributions welcome.
Join 1000+ users for help, tips, discussions, and updates.
Our support team is here to help. Choose your preferred way to reach us:
Monday - Friday: 9AM - 6PM (ICT)
Weekend: Limited support via Discord
Hardware users: Within 24 hours
Managed users: Within 4 hours
English, Khmer, Vietnamese
Choose your path and start chatting with your personal AI assistant today.