OpenClaw is an open-source personal AI assistant platform that lets you interact with AI through multiple messaging channels. You can use Qwen from Alibaba Cloud Model Studio in OpenClaw.
OpenClaw was formerly named Moltbot and Clawdbot. Some commands may not be fully migrated. If you see the command not found: openclaw error when using a new command such as openclaw dashboard, switch to an old command such as clawdbot dashboard or moltbot dashboard.
Model availability
OpenClaw supports text generation models from Model Studio through OpenAI-compatible APIs. The supported models are listed in the following table.
Text generation - Qwen | |
Text generation - Qwen - open source | |
Text generation - third-party models |
If you use OpenClaw frequently or want to reduce costs, subscribe to:
Coding Plan: A fixed monthly fee plan that provides up to 90,000 requests per month, ideal for frequent coding tasks.
Setup steps
1. Activate Model Studio
Create an account: If you do not have an Alibaba Cloud account, register one.
If you encounter any issues during registration, see Create an Alibaba Cloud account.
Activate Model Studio: Use your Alibaba Cloud account to go to the Model Studio console. Read and agree to the Terms of Service. The service is then automatically activated. If you are not prompted to accept the Terms of Service, the service is already activated.
If a message appears indicating that you have not completed identity verification while you are enabling the service, you must first complete identity verification.
After activating Model Studio for the first time, you get a free quota for model inference valid for 90 days. For more information, see Free quota for new users.
You will be charged if you exceed your free quota or if the quota expires. To avoid unexpected charges, enable the Free quota only feature. Final charges are based on the real-time pricing in the console and your final bill.
2. Install OpenClaw
System requirements
OpenClaw requires Node.js 22 or later. Check your Node.js version:
node --versionIf Node.js is not installed or its version is outdated, visit the Node.js official website to download and install it.
Installation methods
macOS / Linux
Use the official installation script:
curl -fsSL https://openclaw.ai/install.sh | bashAlternatively, install it globally using npm:
npm install -g openclaw@latestWindows
In PowerShell, run the following command:
iwr -useb https://openclaw.ai/install.ps1 | iexAlternatively, install it globally using npm:
npm install -g openclaw@latestRun the initial setup wizard
After you install OpenClaw for the first time, the setup wizard starts automatically to help you quickly complete the initial configuration. You can also run the wizard manually by executing the openclaw onboard command.
Configuration item | Recommended setting |
I understand this is powerful and inherently risky. Continue? | Select Yes |
Onboarding mode | Select QuickStart |
Model/auth provider | Select Skip for now (configure Model Studio models later) |
Filter models by provider | Select All providers |
Default model | Select Keep current |
Select channel (QuickStart) | Select Skip for now (configure channel later) |
Configure skills now? (recommended) | Select No |
Enable hooks? | Press the space bar to select. Press Enter to continue. |
How do you want to hatch your bot? | Select Do this later |
Next, configure a Model Studio model for OpenClaw.
3. Configure Model Studio models
To connect OpenClaw to Model Studio, set the following parameters:
Ensure that the Base URL, API key, and model all belong to the same region.
Base URL | By region:
|
API key | Your Model Studio access credential, used for identity verification and billing. Get it from the API Key Management (Beijing), API Key Management (Singapore), or API Key Management (Virginia) page. |
Model ID | The model to use, such as Model support varies by region. For example, the US (Virginia) region does not support |
Disable thinking mode ("reasoning": false) when configuring parameters; otherwise, the response content will be empty.
Configure these settings using one of the following two methods. We recommend the Web console.
Method 1: Web console
OpenClaw includes a web-based editor for the openclaw.json configuration file.
Start the web console
In your terminal, run the following command:
openclaw dashboardYour browser opens the OpenClaw console, typically available at
http://127.0.0.1:18789. If the console does not open automatically, navigate to the URL manually.
Go to the configuration page
In the navigation pane on the left, choose . Then click Raw below the menu.
Add Model Studio configuration
In the navigation pane on the left, choose . Copy the following configuration and use it to replace the original
"agents": {...},section. Then, replace theDASHSCOPE_API_KEYplaceholder with your Model Studio API key.Ensure that the baseUrl and your Model Studio API key belong to the same region. The
reasoningparameter supports onlyfalse."models": { "mode": "merge", "providers": { "bailian": { "baseUrl": "https://dashscope-intl.aliyuncs.com/compatible-mode/v1", "apiKey": "DASHSCOPE_API_KEY", "api": "openai-completions", "models": [ { "id": "qwen3-max-2026-01-23", "name": "qwen3-max-thinking", "reasoning": false, "input": [ "text" ], "cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 262144, "maxTokens": 65536 } ] } } }, "agents": { "defaults": { "model": { "primary": "bailian/qwen3-max-2026-01-23" }, "models": { "bailian/qwen3-max-2026-01-23": { "alias": "qwen3-max-thinking" } }, "maxConcurrent": 4, "subagents": { "maxConcurrent": 8 } } },Save and apply the configuration
Click Save in the upper-right corner. Then click Update.
Method 2: Edit the configuration file
If you prefer to edit the configuration file directly, follow these steps. The file is located at ~/.openclaw/openclaw.json and is automatically read by OpenClaw at startup.
Open the configuration file
Open the file in a text editor. If the file does not exist, the editor will create it.
# You can use any editor, such as vim. # vim ~/.openclaw/openclaw.json nano ~/.openclaw/openclaw.jsonAdd Model Studio configuration
Copy and paste the following configuration into the file. Replace the
DASHSCOPE_API_KEYplaceholder with your Model Studio API key.The
reasoningparameter supports onlyfalse.{ "meta": { "lastTouchedVersion": "2026.2.1", "lastTouchedAt": "2026-02-03T08:20:00.000Z" }, "models": { "mode": "merge", "providers": { "bailian": { "baseUrl": "https://dashscope-intl.aliyuncs.com/compatible-mode/v1", "apiKey": "DASHSCOPE_API_KEY", "api": "openai-completions", "models": [ { "id": "qwen3-max-2026-01-23", "name": "qwen3-max-2026-01-23", "reasoning": false, "input": ["text"], "contextWindow": 262144, "maxTokens": 65536 }, { "id": "qwen3-coder-plus", "name": "qwen3-coder-plus", "reasoning": false, "input": ["text"], "contextWindow": 131072, "maxTokens": 32768 } ] } } }, "agents": { "defaults": { "model": { "primary": "bailian/qwen3-max-2026-01-23" } } }, "gateway": { "mode": "local", "auth": { "mode": "token", "token": "test123" } } }Save the configuration file
If you use nano:
Press
Ctrl+Xto exit.Press
Yto confirm that you want to save the changes.Press
Enterto confirm the filename.
If you use vim:
Press
Escto exit insert mode.Type
:wqand pressEnterto save the file and exit.
Verify the configuration
View the configuration file and confirm that the configuration is correct:
cat ~/.openclaw/openclaw.json
4. Start chatting
Command line
In your Elastic Compute Service (ECS) instance or local terminal, start a chat session in the terminal:
openclaw tui
Web interface
In your ECS instance or local terminal, start a chat session in your browser:
openclaw dashboard
FAQ
Model not found or response is empty
Check the following items:
Ensure that the model ID is spelled correctly.
Ensure that the
providername in your configuration file matches the reference in your command. For example, if your configuration file usesbailianas the provider, use a reference such asbailian/qwen3-max-2026-01-23.Set the
reasoningparameter tofalse.
“command not found” when running openclaw
This error may occur for the following reasons:
The OpenClaw installation was not successful. To verify the installation, run the
openclaw --versioncommand in your ECS instance or local terminal. If a version number is returned, the installation was successful.Not all commands were fully migrated during the OpenClaw renaming process. Try the following:
Use the old commands, such as
moltbotorclawdbot.Reinstall the latest version by running the following command:
npm install -g openclaw@latest.