OpenClaw Integration Guide for the Turing Platform

OpenClaw is an open-source personal AI assistant platform that supports interacting with AI through multiple messaging channels. By configuring it appropriately, you can access the powerful models available on the Turing Platform.

About OpenClaw

OpenClaw was renamed from Moltbot / Clawdbot. The current package name is openclaw. If the openclaw command is unavailable, the installation likely did not complete successfully or your PATH has not been refreshed — reopen your terminal, or reinstall/upgrade with npm install -g openclaw@latest. clawdbot and moltbot are deprecated legacy packages that are no longer updated with this project. Do not use them to install the current version.

Prerequisites

Before getting started, make sure you have:

• A Turing Platform API key
• A supported Node.js version installed: 22.22.3+, 24.15+, or 25.9+ (Node 24 recommended; note that Node 23 is not supported)

Quick Start

1. Check Your Node.js Version

OpenClaw requires Node.js 22.22.3+, 24.15+, or 25.9+ (Node 24 recommended; Node 23 is not supported). Check your version with:

node --version

If Node.js is not installed or the version is too old, visit the Node.js website to download and install it.

2. Install OpenClaw

macOS / Linux

The official install script is recommended:

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

Or install globally via npm:

npm install -g openclaw@latest

Windows

Install globally via npm:

npm install -g openclaw@latest

After installation, verify it:

openclaw --version

Command Compatibility

The current command name is openclaw. If it is unavailable, reopen your terminal to refresh the PATH, or reinstall/upgrade with npm install -g openclaw@latest — do not fall back to the deprecated clawdbot / moltbot legacy packages.

3. Initialize Configuration

After installation, run the onboarding wizard:

openclaw onboard

Follow the recommendations below when going through the wizard:

Setting	Example Selection	Description
I understand this is powerful and inherently risky. Continue?	Yes	Confirm to continue
Onboarding mode	QuickStart	Quick configuration mode
Model/auth provider	Custom Provider	Select a custom provider
API Base URL	https://live-turing.cn.llm.tcljd.com/api/v1	Turing Platform API endpoint
How do you want to provide this API key?	Paste API key now	Paste your API key directly
API Key	Enter your Turing Platform API key	The key obtained from the Turing Platform
Endpoint compatibility	OpenAI-compatible	Select OpenAI-compatible mode
Model ID	minimax-m3	The model ID to use; you can start with the example model to verify, and see the model list for more available models
Endpoint ID	eaglelab	A unique identifier for the custom provider. You can use any name. This name becomes the key under models.providers and is used as the prefix in the model path in agents configuration (see below)
Model alias (optional)	minimax-m3 (optional)	An alias for the model for easier identification
Filter models by provider	All providers	Show models from all providers
Default model	Keep current	Keep the current selection
Select channel (QuickStart)	Skip for now	Configure channels later
Configure skills now?	No	Configure skills later
Enable hooks?	Press Space to select desired options	Choose as needed
How do you want to hatch your bot?	Do this later	Configure later

Relationship Between Endpoint ID and Model Path

The Endpoint ID entered in the onboard wizard is the provider key under models.providers. In the agents configuration, the model path format is:

{Endpoint ID}/{Model ID}

Example:

Endpoint ID	Model ID	Model path in agents
eaglelab	minimax-m3	eaglelab/minimax-m3

Therefore, in the configuration examples below, both "primary" and the keys in "models" within the "agents" section must use the full {Endpoint ID}/{Model ID} format.

Important

• Keep your API key secure and do not share it with others.
• The API key will be stored in the local configuration file (~/.openclaw/openclaw.json).

Configuring Turing Platform Models

OpenClaw offers two configuration methods. Choose whichever suits you best.

Method 1: Web Console Configuration

OpenClaw provides a web-based editor for modifying the openclaw.json configuration file.

1. Start the Web Console

Run the following command in your terminal:

openclaw dashboard

The browser will open the OpenClaw console automatically (usually at http://127.0.0.1:18789). If it does not open automatically, navigate to that address manually.

2. Open the Configuration Page

In the left sidebar, navigate to: Config > Authentication, then click Raw at the bottom of the menu.

3. Add the Turing Platform Configuration

In the configuration editor, locate the "agents": {...} section, replace the "models" and "agents" sections, and replace YOUR_TURING_API_KEY with your Turing Platform API key.

About Endpoint ID

The key in "models": { "providers": { ... } } (in the example below, "eaglelab") is the Endpoint ID. In the "agents" section, the model path must be written as {Endpoint ID}/{Model ID}.

• In this example, the Endpoint ID is eaglelab and the Model ID is minimax-m3, so the path is eaglelab/minimax-m3.
• If your Endpoint ID is my-provider and the Model ID is minimax-m3, the path is my-provider/minimax-m3.

  "models": {
    "mode": "merge",
    "providers": {
      "eaglelab": {                        // ← This key is the Endpoint ID; it can be customized
        "baseUrl": "https://live-turing.cn.llm.tcljd.com/api/v1",
        "apiKey": "YOUR_TURING_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "minimax-m3",
            "name": "MiniMax M3",
            "input": ["text", "image"],
            "contextWindow": 1048576,
            "maxTokens": 32768
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "eaglelab/minimax-m3"     // ← Format: {Endpoint ID}/{Model ID}
      },
      "models": {
        "eaglelab/minimax-m3": {             // ← Same as above; use {Endpoint ID}/{Model ID} as the key
          "alias": "minimax-m3"
        }
      },
      "maxConcurrent": 4,
      "subagents": {
        "maxConcurrent": 8
      }
    }
  },

4. Save and Apply the Configuration

After completing the configuration, click Save in the top-right corner, then click Update to apply the changes.

Method 2: Manually Edit the Configuration File

If you prefer to edit the configuration file directly, follow these steps. The configuration file is located at ~/.openclaw/openclaw.json and is loaded automatically when OpenClaw starts.

1. Open the Configuration File

Open the configuration file with a text editor. If the file does not exist, the editor will create it:

# Using the nano editor
nano ~/.openclaw/openclaw.json

# Or using the vim editor
# vim ~/.openclaw/openclaw.json

2. Add the Turing Platform Configuration

In the configuration file, locate the "agents": {...} section, replace the "models" and "agents" sections, and replace YOUR_TURING_API_KEY with your Turing Platform API key.

About Endpoint ID

The key in "models": { "providers": { ... } } (in the example below, "eaglelab") is the Endpoint ID. In the "agents" section, the model path must be written as {Endpoint ID}/{Model ID}.

• In this example, the Endpoint ID is eaglelab and the Model ID is minimax-m3, so the path is eaglelab/minimax-m3.
• If your Endpoint ID is my-provider and the Model ID is minimax-m3, the path is my-provider/minimax-m3.

  "models": {
    "mode": "merge",
    "providers": {
      "eaglelab": {                        // ← This key is the Endpoint ID; it can be customized
        "baseUrl": "https://live-turing.cn.llm.tcljd.com/api/v1",
        "apiKey": "YOUR_TURING_API_KEY",
        "api": "openai-completions",
        "models": [
          {
            "id": "minimax-m3",
            "name": "MiniMax M3",
            "input": ["text", "image"],
            "contextWindow": 1048576,
            "maxTokens": 32768
          }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "eaglelab/minimax-m3"     // ← Format: {Endpoint ID}/{Model ID}
      },
      "models": {
        "eaglelab/minimax-m3": {             // ← Same as above; use {Endpoint ID}/{Model ID} as the key
          "alias": "minimax-m3"
        }
      },
      "maxConcurrent": 4,
      "subagents": {
        "maxConcurrent": 8
      }
    }
  },

3. Save the Configuration File

If using the nano editor:

1. Press Ctrl+X to exit the editor.
2. Press Y to confirm saving when prompted.
3. Press Enter to confirm the file name.

If using the vim editor:

1. Press Esc to ensure you are out of insert mode.
2. Type :wq and press Enter to save and exit.

4. Verify the Configuration

Use the following command to view the configuration file and confirm it is correct:

cat ~/.openclaw/openclaw.json

Configuration Reference

Setting	Description
baseUrl	Turing Platform API endpoint: https://live-turing.cn.llm.tcljd.com/api/v1
apiKey	Turing Platform API key, obtained from the Turing Platform Portal
api	API type; use openai-completions for the OpenAI-compatible interface
models	List of available models, including model ID, name, input types, context window, and other details; refer to the model list for valid model IDs
agents.defaults.model.primary	The default primary model, in the format {Endpoint ID}/{Model ID}, where Endpoint ID is the provider key defined under models.providers, e.g., eaglelab/minimax-m3
gateway.mode	Gateway mode; local indicates local mode
gateway.auth	Authentication configuration for Web console access

Recommended Models

OpenClaw uses the OpenAI-compatible interface. Set the Model ID to the interface model ID from the model list. The example configuration uses minimax-m3, but you can choose based on provider capabilities and pricing:

Model Selection Recommendations

• MiniMax: Suited for general-purpose agents and long-context tasks.
• Alibaba Qwen: Suited for general conversation, tool use, and multimodal scenarios.
• Zhipu GLM: Suited for complex reasoning and long-document tasks.
• OpenAI / Claude: Suited for coding, reasoning, and high-quality text tasks.

Start Chatting

Once configured, you can use OpenClaw in the following ways:

Terminal UI (TUI)

Launch the interactive interface in your terminal:

openclaw tui

This is the most commonly used interactive interface, supporting real-time conversation and command execution.

Web Console

Launch the web interface for configuration and conversation:

openclaw dashboard

The browser will automatically open the console (usually at http://127.0.0.1:18789). In the web interface you can:

• Have AI conversations
• Switch between different models
• Modify configuration parameters
• View conversation history

Connect to Chat Tools

OpenClaw supports multi-channel interaction and can integrate with chat tools such as Feishu and Discord, embedding your AI assistant into your daily workflow for 24/7 continuous service. For detailed configuration, refer to the OpenClaw official documentation.

Frequently Asked Questions

Issue 1: Model Not Found or Empty Responses

Please verify that:

• The model ID is spelled correctly.
• The API key is valid and has sufficient quota.
• Your network connection is working and can reach the Turing Platform API.

Issue 2: "command not found" When Running the openclaw Command

Possible causes and solutions:

1. OpenClaw installation did not complete: Run openclaw --version; if a version number is printed, the installation was successful.
2. PATH has not been refreshed: Reopen your terminal, or confirm that the global npm bin directory is in your PATH.
3. Reinstallation required: Run npm install -g openclaw@latest to install the latest version.

Issue 3: Where Is the Configuration File?

OpenClaw's files are located at:

• Configuration file: ~/.openclaw/openclaw.json
• Log files: ~/.openclaw/logs/

You can edit the configuration file directly to change settings.

Tips and Tricks

1. Switch Models in the Web Console

Start the Web console:

openclaw dashboard

In the console interface, you can:

• Switch between different models in the settings page.
• View detailed model information and capabilities.
• Configure conversation parameters (temperature, max tokens, etc.).

2. Quick Conversations via the Command Line

If you just need to ask a quick question, use TUI mode:

openclaw tui

3. Integrate with Chat Tools

OpenClaw supports integration with tools such as DingTalk and WeCom, giving your AI assistant 24/7 availability. For detailed configuration, refer to the OpenClaw official documentation.

Resources

• OpenClaw Official Website
• OpenClaw GitHub
• OpenClaw Official Documentation
• Turing Platform Model List
• Enterprise Access Application

Technical Support

Having trouble? Visit Technical Support & Contact for help.

Back

← Back to AI Coding Tools Overview