Singularity.Kiwi AI progress, with honesty.
← Back to Toolkit

🦞 Set Up OpenClaw

Connect your chat apps to AI agents. 15 minutes, beginner-friendly. Free to start.

🕐Time15–30 min
📊DifficultyBeginner
💰Cost$0–20/mo
💻PlatformMac · Win · Linux

OpenClaw is free, open-source, and genuinely practical. It connects your chat apps to AI agents through a Gateway that runs on your computer. openclaw.ai

💡 Prefer a desktop app with built-in memory? Try Hermes Agent instead → — same concept, native Mac app, persistent memory out of the box.

Step 1: Install Ollama & Get an API Key

OpenClaw needs an AI model to think with. The easiest way is Ollama Cloud — sign up free, grab an API key.

  1. Go to ollama.com and download the app for your OS
  2. Install it (about a minute) and run the app once
  3. Sign up for a free account → Settings → API Keys → Create API Key
  4. Copy the key and save it somewhere safe — Ollama only shows it once

Ollama Cloud pricing:

  • Free ($0) — Testing & light use. Hits limits during longer tasks.
  • Pro ($20/mo) — 50x more usage, larger models, run 3 at once. Sweet spot for daily use.
  • Max ($100/mo) — 5x Pro, 10 models at once. Only for heavy sustained work.

💡 Recommendation: Start on Free to verify the install, then upgrade to Pro. $20/mo is enough for most users.

Don't want Ollama? You can also use OpenAI or Anthropic (both pay-as-you-go, no free tier).

Step 2: Install OpenClaw

🍎 macOS
🪟 Windows
🐧 Linux
1

Option A: Download the Mac App

Make sure the Ollama app is installed and running (Step 1 above).

Download from openclaw.ai

Open the app and follow the setup prompts. Sign in with your Ollama API key when asked. Pick a model. Choose Telegram, Discord, or just use the built-in chat.

Requires macOS 15+. No terminal needed.

2

Option B: One-Line Terminal Install

Make sure the Ollama app is running, then paste:

curl -fsSL https://openclaw.ai/install.sh | bash

This installs Node.js and OpenClaw automatically. When it finishes, run the setup wizard:

openclaw onboard

The wizard asks:

  • Quick start or manual? — Pick Quick start
  • AI provider — Pick Ollama Cloud and paste your API key
  • Model — Pick one from the list
  • Telegram — Skip for now if you don't have a bot token
  • Auto-start — Say yes

After the wizard, the Gateway starts automatically.

3

Grant Permissions (App Only)

If using the desktop app, OpenClaw will request:

  • Accessibility: Required for keyboard shortcuts and automation
  • Full Disk Access: Required to read/write files in your folders
  • Automation: Required to control other apps

Go to System Preferences → Privacy & Security and enable these for OpenClaw.

1

Download OpenClaw

Download the Windows installer:

Download from openclaw.ai GitHub Releases
2

Run the Installer

  1. Run the downloaded .exe file
  2. Follow the installation wizard
  3. Choose install location (default is fine)
  4. Launch OpenClaw from the Start Menu

Windows may show a "Windows protected your PC" warning. Click "More info" then "Run anyway" if you downloaded from the official source.

3

Grant Permissions

OpenClaw may request:

  • File system access: To read and write files
  • Network access: To connect to AI services
  • Windows Terminal: For running commands

Approve these in the Windows security dialogs that appear.

4

Install Ollama (for local models)

  1. Download Ollama from ollama.com
  2. Run the installer
  3. Open PowerShell and run: ollama pull qwen3-coder
5

Configure & Start

  1. Open OpenClaw from Start Menu
  2. Choose your default model
  3. Set up preferences
  4. Start chatting!
1

Download OpenClaw

Download the Linux package (AppImage or .deb/.rpm):

Download from openclaw.ai GitHub Releases
2

Install

For AppImage:

chmod +x openclaw-*.AppImage
./openclaw-*.AppImage

For Debian/Ubuntu (.deb):

sudo dpkg -i openclaw-*.deb
sudo apt-get install -f  # Fix dependencies

For Fedora/RHEL (.rpm):

sudo rpm -i openclaw-*.rpm
3

Install Ollama

curl -fsSL https://ollama.com/install.sh | sh
ollama pull qwen3-coder
4

Run OpenClaw

Launch from your application menu or run:

openclaw

Step 3: Test & Switch Models

Test it

Open the OpenClaw app or chat window and try:

  • Say "Hello"
  • Ask "What's 2+2?" — should say 4
  • Say "My name is Chris", then ask "What's my name?"

If all three work, you're set up.

Switch models

With the Ollama app running and your API key set up, the fastest way to try a different model:

ollama launch openclaw --model glm-5.2:cloud

This downloads the model if you don't have it yet and reconnects OpenClaw in one step. Browse available models at ollama.com/library.

Recommended models (June 2026):

  • glm-5.2:cloud — Excellent for writing, publishing, research, and general agent work. Our current pick.
  • ollama/qwen3.5:cloud — Fast, cheap, capable
  • openai/gpt-4o — Excellent reasoning and writing (requires OpenAI key)
  • anthropic/claude-3.5-sonnet — Careful, great for analysis (requires Anthropic key)

💡 Cost tip: Start on Ollama Cloud at $20/month. Only move to OpenAI or Anthropic if you need noticeably better quality for a specific task. See AI Pricing Guide for full comparisons.

🛠️ Other Ways to Run Local AI

LM Studio

  • GUI app for downloading and chatting with models
  • Browse HuggingFace models inside the app
  • Easiest for beginners — no terminal needed
  • lmstudio.ai

OpenClaw + Ollama

  • AI assistant with memory, tools, and integrations
  • Telegram, Discord, iMessage, WhatsApp
  • File access, terminal, web search, automation
  • openclaw.ai

Ollama (CLI Only)

  • Terminal-based model runner
  • Lightweight, no GUI needed
  • Also exposes OpenAI-compatible API on port 11434
  • ollama.com

vLLM / SGLang

  • Production inference servers
  • For serving models to multiple users
  • Requires NVIDIA GPU or ROCm
  • Best for teams and self-hosted APIs

⚙️ Initial Configuration

After installation, OpenClaw will guide you through setup. Here's what to configure:

Model Selection

  • Local (Ollama): Free, private, offline-capable. Try qwen3-coder for coding or deepseek-r1:14b for reasoning.
  • Cloud: More powerful but sends data to servers. Add API keys for ChatGPT, Claude, etc.

Memory Setup

OpenClaw can remember information across conversations. In settings:

  • Enable "Long-term memory" to remember facts about you
  • Set memory location (default is fine)
  • Configure what to remember/forget

Integrations (Optional)

Connect to messaging platforms in Settings → Integrations:

  • iMessage: Chat with AI via Messages app (Mac only)
  • Telegram: Create a bot for Telegram chats
  • Discord: Add AI to your Discord server

🔧 Troubleshooting

"Ollama not found" error

Make sure Ollama is installed and running:

# Check if Ollama is running
ollama list

# If not, start it
ollama serve

"Permission denied" errors

On macOS, make sure OpenClaw has:

  • Accessibility access (System Preferences → Privacy & Security)
  • Full Disk Access (System Preferences → Privacy & Security)

Slow responses

  • Use a smaller model (phi4-mini or llama3.2:3b for 8GB RAM)
  • Close other resource-intensive apps
  • Check you have enough RAM for the model size

Need more help?

Documentation GitHub Issues Discord Community

📖 Want More Help?

The Lobster Farm is our free ebook that walks you through building your own local AI agent stack — the same approach we used to build Singularity.Kiwi. 12 chapters, practical, beginner-friendly.

Get The Lobster Farm (Free) →