Getting Started with Pico Claw: Quick Start Guide

P/N: AIBOT-CLAW-HOSTING

Cloud Infrastructure Module

Your ultra-lightweight bot deserves professional cloud infrastructure. Deploy to managed hosting with zero maintenance overhead.

24/7
UPTIME
🛡️
SECURE
INFRA
Deployment <60s
Availability 99.9%
Auto Updates ENABLED
Monitoring 24/7
Package Cost
$16.99
per month • annual billing
Deploy Module

Get Pico Claw up and running in minutes. This quick start guide walks you through hardware selection, installation, configuration, and your first interaction with picoclaw. From zero to AI assistant in under 30 minutes.

Pico Claw Quick Start Overview

This guide covers everything you need to get Pico Claw running. Follow these steps in order for a smooth setup experience. Total time: approximately 20-30 minutes.

What You'll Accomplish

  • Choose and set up hardware
  • Install Pico Claw binary
  • Configure API keys and platforms
  • Run PicoClaw and verify it works
  • Send your first AI query
⏱️

Time Required

  • Hardware prep: 5-10 minutes
  • Installation: 5 minutes
  • Configuration: 10 minutes
  • Testing: 5 minutes
  • Total: ~20-30 minutes
📋

Prerequisites

  • Linux device (RISC-V/ARM/x86)
  • 64MB+ RAM (256MB+ recommended)
  • Network connectivity
  • AI API key (OpenAI/Anthropic)
  • Basic command line knowledge

Choose Hardware for Pico Claw

Pico Claw runs on diverse hardware platforms. Select based on your budget, use case, and availability. All options provide excellent picoclaw performance.

LicheeRV Nano

$9.9 - Recommended

Best value for Pico Claw. Native RISC-V support, 256MB RAM, ultra-compact form factor. Perfect for beginners and edge deployments. Boots picoclaw in under 1 second.

  • SOPHGO SG2002 RISC-V @ 1GHz
  • 256MB DDR3 (25x PicoClaw needs)
  • Ethernet or WiFi 6 variants
  • Only $9.9 USD
Setup Guide
🍓

Raspberry Pi Zero

~$15 - Popular Choice

Excellent if you already have Pi hardware. 512MB RAM provides massive headroom for Pico Claw. Large community and extensive documentation available.

  • ARM 64 fully supported
  • 512MB RAM
  • Boots PicoClaw in <2 seconds
  • Familiar ecosystem
🖥️

NanoKVM

$30-50 - Server Management

Combines IP-KVM functionality with Pico Claw. Perfect for data center automation and remote server management. Based on LicheeRV Nano core.

  • IP-KVM + AI assistant
  • HDMI capture included
  • Network management tools
  • Enterprise use cases
💻

Any Linux Device

Universal Compatibility

Pico Claw runs on any Linux system with 64MB+ RAM. Desktop, server, VM, or embedded. RISC-V, ARM, and x86 architectures all supported.

  • x86_64, ARM64, RISC-V 64
  • Minimum 64MB RAM
  • Any Linux distribution
  • Virtual machines supported

Prepare for Pico Claw Installation

Before installing Pico Claw, complete these preparation steps. Having everything ready ensures a smooth installation process.

Preparation Checklist

1. Set Up Hardware

Ensure your chosen hardware is powered on, running Linux, and accessible via SSH or direct terminal. Update system packages to latest versions.

ssh user@your-device
sudo apt update && sudo apt upgrade -y
2. Configure Network

Verify network connectivity. PicoClaw needs internet access to reach AI APIs and platform services (Telegram, Discord, etc.).

ping -c 3 google.com
# Should show successful responses
3. Obtain AI API Key

Sign up for OpenAI, Anthropic Claude, or another supported AI service. Generate API key from their dashboard. Keep it secure.

Required for Pico Claw to function.
Store securely - never commit to git.
4. Create Platform Bots (Optional)

If using Telegram: create bot via @BotFather. If using Discord: create application in Discord Developer Portal. Get bot tokens for config.yaml.

Telegram: @BotFather for bot token
Discord: discord.com/developers

Pico Claw Installation Methods

Choose your preferred installation method. Pre-built binary is fastest and recommended for most users. Build from source if you need customization.

Recommended

Quick Install (Pre-built Binary)

Fastest method - download and run in seconds.

Download for Your Architecture

# RISC-V 64
wget https://github.com/sipeed/picoclaw/releases/latest/download/picoclaw-riscv64

# ARM 64
wget https://github.com/sipeed/picoclaw/releases/latest/download/picoclaw-arm64

# x86_64
wget https://github.com/sipeed/picoclaw/releases/latest/download/picoclaw-x86_64

Make Executable

chmod +x picoclaw-*
mv picoclaw-* picoclaw

Build from Source

For developers who want to customize Pico Claw.

Install Go (if not installed)

wget https://go.dev/dl/go1.21.linux-$(uname -m).tar.gz
sudo tar -C /usr/local -xzf go1.21.linux-*.tar.gz
export PATH=$PATH:/usr/local/go/bin

Clone and Build

git clone https://github.com/sipeed/picoclaw.git
cd picoclaw
go build -o picoclaw

Configure Pico Claw for First Use

Configure Pico Claw with your API keys and platform credentials. The config.yaml file controls all picoclaw settings and connections.

⚙️

Initialize Configuration

Run initialization to create default config.yaml. This generates a template with all configuration options explained.

./picoclaw --init
# Creates config.yaml in current directory
🔑

Add AI API Key

Edit config.yaml and add your AI service API key. This is required for Pico Claw to function. Use OpenAI, Anthropic Claude, or compatible service.

nano config.yaml
# Set: api_key: "your-api-key-here"
💬

Configure Platforms

Add bot tokens for messaging platforms you want to use. Pico Claw supports Telegram (full), Discord (full), QQ, DingTalk, Feishu, and WhatsApp.

Telegram: telegram_token: "bot-token"
Discord: discord_token: "bot-token"
Configuration Tips:
  • Start with just AI API key for testing
  • Add platforms incrementally
  • Keep config.yaml secure (contains secrets)
  • See detailed configuration guide for all options

Running Pico Claw for the First Time

Start Pico Claw and verify everything works correctly. Your first run should complete in seconds with PicoClaw connecting to all configured services.

First Run Steps

1. Start Pico Claw

Run the picoclaw binary. It should boot in under 1 second and display connection status.

./picoclaw
# Watch for successful startup messages
2. Verify Connections

Check logs to confirm PicoClaw connected to AI service and platforms successfully.

Look for: "Connected to OpenAI"
"Telegram bot started"
"Discord bot ready"
3. Send Test Message

Send "hello" or "ping" through your configured platform. Pico Claw should respond within seconds.

Via Telegram: message your bot
Via Discord: @mention your bot
Should see AI-generated response
4. Monitor Performance

Check memory usage and response times. Should see <10MB RAM usage and fast responses.

top -p $(pgrep picoclaw)
# Monitor resource usage
Common Issues:
  • Connection Failed: Check network connectivity and API key validity
  • Bot Not Responding: Verify bot token is correct and bot is added to channel
  • High Memory: Normal on first run; should stabilize under 10MB
  • Permission Denied: Ensure binary is executable (chmod +x)

Next Steps with Pico Claw

Now that Pico Claw is running, explore advanced features and customization options. Make picoclaw your own with skills, multi-platform support, and community tools.

Add More Platforms

Connect Pico Claw to multiple messaging platforms simultaneously. Run one picoclaw instance across Telegram, Discord, QQ, DingTalk, and more. Unified AI assistant everywhere you communicate.

🛠️

Install Skills

Extend Pico Claw capabilities with skills from GitHub. Add tools for web search, code execution, file operations, and more. Install with simple git clone commands.

⚙️

Advanced Configuration

Explore config.yaml options for memory settings, persistent storage, custom prompts, and platform-specific features. Fine-tune picoclaw for your specific needs.

👥

Join Community

Connect with other Pico Claw users. Share skills, configurations, and use cases. Get help, report issues, and contribute to the growing picoclaw ecosystem.

Helpful Resources:

Your Pico Claw Journey Starts Here

You now have everything you need to run PicoClaw. Start with $9.9 hardware or any Linux device. Join thousands using ultra-lightweight AI on the edge.