CherryIN Docs

CherryIN now supports Kimi K2.5! OpenClaw (formerly ClawdBot) has been trending worldwide with over 120K stars!

This tutorial will guide you step-by-step to configure OpenClaw + CherryIN and build your own AI Agent!

Before You Start

What is OpenClaw?

OpenClaw is an open-source AI Agent framework that allows AI models to control your computer for various tasks—writing code, managing files, executing commands, and more.

Why Use a Virtual Machine?

Since OpenClaw can directly control your computer, it's safer to use it in a virtual machine. If anything goes wrong, your main system stays safe.

Checklist

⬜ A computer (Windows / macOS / Linux)
⬜ Stable network connection
⬜ About 20 minutes
⬜ (Recommended) A virtual machine environment

Step 1: Install OpenClaw

While the demo uses Linux, the process is cross-platform!

Windows Users

Windows users have two installation options. WSL is recommended:

📦 Option 1: WSL (Recommended)

WSL (Windows Subsystem for Linux) lets you run a Linux environment on Windows for a smoother experience.

1. Install WSL

Open PowerShell as Administrator and run:

wsl --install

Restart your computer after installation.

2. Set Up Linux User

After restart, Ubuntu terminal will open automatically. Follow the prompts to set your username and password.

3. Continue

Once set up, follow the Linux steps below in your WSL terminal.

🪟 Option 2: Native Windows Installation

If you prefer not to use WSL, you can install directly on Windows.

1. Install Node.js

Visit Node.js website, download the LTS version (22 or above recommended), and run the installer.

2. Verify Installation

Open PowerShell or CMD and run:

node -v
npm -v

You should see version numbers.

3. Install OpenClaw

npm install -g openclaw@latest

1.1 Install Node.js (via nvm)

For macOS / Linux users, we recommend using nvm to install Node.js:

# Download and install nvm
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.40.3/install.sh | bash

# Load nvm (avoid restarting terminal)
\. "$HOME/.nvm/nvm.sh"

# Install Node.js 24 (version 22+ recommended)
nvm install 24

# Verify installation
node -v  # Should show v24.x.x
npm -v   # Should show 11.x.x

1.2 Install OpenClaw

With Node.js ready, install OpenClaw with one command:

npm install -g openclaw@latest
# Or with pnpm: pnpm add -g openclaw@latest

✅ Success Check

Run these commands—if you see version numbers, installation is complete:

node -v
openclaw --version

Step 2: Initialize OpenClaw

Run the initialization command:

openclaw onboard --install-daemon

The OpenClaw configuration page will open automatically. If not, run openclaw onboard manually.

Accept Security Notice

Since OpenClaw can control your computer, there's a security notice. Accept it to continue.

Skip Provider Setup

Skip the Provider setup for now. We'll configure CherryIN through the config file later.

Skip Notification Settings

These IM notifications require additional setup. Skip them for now—we'll get the basics running first.

Select WebUI Mode

Choose WebUI mode for browser-based interaction.

✅ Success Check

When you see the console output with the access URL, initialization is complete:

Step 3: Register for CherryIN and Get API Key

3.1 Create Account

Visit CherryIN: https://open.cherryin.ai/

Email Login Recommended

We strongly recommend using email login for better stability. GitHub login may occasionally have issues.

3.2 Add Credits

After adding credits, you can access all models including free DeepSeek v3.2 and the latest models.

3.3 Create Token

Go to "Token Management" and create a new token. Name it something like Openclaw-Agent for easy identification.

Copy and save your token—you'll need it for configuration.

✅ Success Check

You now have a token starting with sk-, like sk-xxxxxxxxxxxxxxxx.

Step 4: Configure Model Access

4.1 Locate Config File

OpenClaw config file path:

~/.openclaw/openclaw.json

Note

~ refers to your home directory. On Linux it's /home/your-username, on macOS it's /Users/your-username.

4.2 Add CherryIN Configuration

Open the config file with your preferred editor and add the following. Replace sk-xxxxx with your actual token:

{
  "models": {
    "mode": "merge",
    "providers": {
      "cherryin": {
        "baseUrl": "https://open.cherryin.cc/v1",
        "apiKey": "sk-xxxxx",
        "auth": "api-key",
        "api": "openai-completions",
        "headers": {},
        "models": [
          {
            "id": "agent/kimi-k2.5",
            "name": "Kimi K2.5",
            "api": "openai-completions"
          },
          {
            "id": "anthropic/claude-opus-4.5",
            "name": "Claude Opus 4.5",
            "api": "openai-completions"
          },
          {
            "id": "anthropic/claude-sonnet-4.5",
            "name": "Claude Sonnet 4.5",
            "api": "openai-completions"
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "cherryin/agent/kimi-k2.5"
      },
      "models": {
        "cherryin/agent/kimi-k2.5": {},
        "cherryin/anthropic/claude-opus-4.5": {},
        "cherryin/anthropic/claude-sonnet-4.5": {}
      },
      "compaction": {
        "mode": "safeguard"
      },
      "maxConcurrent": 4,
      "subagents": {
        "maxConcurrent": 8
      }
    }
  }
}

Find Model IDs

Visit CherryIN Pricing Page to see all available model IDs.

/new

5.2 Switch Model

/model cherryin/agent/kimi-k2.5

Model Path Format

OpenClaw uses the format Provider/ModelID. Our provider ID is cherryin, model ID is agent/kimi-k2.5, so the full path is cherryin/agent/kimi-k2.5.