Simple Application Server:OpenClaw (formerly MoltBot, Clawbot) FAQ

What’s new in the OpenClaw 2026.2.3 application image compared to Moltbot 2026.1.27-beta.1?

• The built-in software is upgraded to OpenClaw and supports all features released by the community before February 3, 2026.

• Web search using the built-in SearXNG Skill is enabled by default in all regions. No additional configuration or fees are required.

• Supports coordinated text processing and image recognition. The default image model is qwen3-vl-plus.

• OpenClaw 2026.2.3 includes Docker to support running agents in sandboxed Docker environments.

How do I update my application image to OpenClaw 2026.2.3? How do I deploy the OpenClaw image on an existing Simple Application Server?

1. Log on to the Simple Application Server console. Select your instance and click Reset System.

2. Select Reset to Other Image and choose OpenClaw 2026.2.3.

3. Reconfigure OpenClaw. After the reset, your previous API key and token are invalid. Go to Application Details to reconfigure the API key.

How do I reset my current application image to the latest version?

If your current image is Moltbot 2026.1.27-beta.1 and you want to try the latest OpenClaw features, see How do I update my application image to OpenClaw 2026.2.3? How do I deploy the OpenClaw image on an existing Simple Application Server? to change the image.

1. Log on to the Simple Application Server console. Select your instance and click Reset System.

2. Select Reset Operating System.

3. Reconfigure OpenClaw. After the reset, your previous API key and token are invalid. Go to Application Details to reconfigure the API key.

Can OpenClaw, deployed on a Simple Application Server, control applications on my local computer?

Not supported. Because the Simple Application Server runs in an isolated cloud network environment, it cannot directly control applications on your local desktop across networks.

Why can’t I use web search even after updating the OpenClaw application image to the latest version?

Go to the Simple Application Server console and check the Image Information section. Confirm that your image has been updated to OpenClaw 2026.2.3. If not, follow How do I update my application image to OpenClaw 2026.2.3? How do I deploy the OpenClaw image on an existing Simple Application Server? to update. This version enables web search using the built-in SearXNG Skill in all regions. No extra setup or fees are required. You can tell OpenClaw to use the SearXNG Skill for web search. If your server is in Hong Kong (China) or outside China, see How do I configure Brave Search for web search in OpenClaw? to set up Brave Search API.

How do I configure other model service providers in OpenClaw? How do I switch to a third-party model service?

The current OpenClaw image comes with the Alibaba Cloud Model Studio (Bailian) Provider pre-initialized. You can directly modify the specific model it uses in the settings. To connect to model services from OpenAI, Anthropic, or Google Gemini, follow these steps.

1. Log on to the server terminal and switch to the root user.

   sudo -i

2. Change the model provider (using OpenRouter as an example).

   openclaw onboard

   

3. Verify the changes. Log on to the web console and check the model configuration to confirm that the new model appears.

   

How do I change the model that OpenClaw calls?

1. Go to the Simple Application Server console on the Servers page. On the card for the server where OpenClaw is deployed, click the instance ID to go to the Server Overview page.

2. Click the Application Details tab. On the Access Control page, click Quick View to go to the Clawdbot conversation page.

3. Click Config > Agents, search for Primary Model or Image Model, and change it to a supported model from Model Studio. Use the format alibaba-cloud/ModelCode, such as alibaba-cloud/qwen3-max-2026-01-23. Find model codes on the Model Studio Model Square page.

   

How do I install or add Skills in OpenClaw?

OpenClaw supports adding a skill using three methods: conversational interaction, link-based installation, and copy-based installation. The steps for each method are as follows:

• Interactive creation (for new Skills).

  OpenClaw has a built-in Skill Creator component. You can chat directly with the Skill Creator, describe your requirements, and let it automatically create a new Skill for you.

• Install from a URL (for existing Skills).

  To install a Skill that someone else has already developed, send the Skill's URL to OpenClaw, and the system will automatically complete the installation. For example, to install the Web Artifacts Builder provided by Anthropic, tell the Agent to install this Skill.

• Install from a transit source (for networks with restricted access).

  If your server's network environment cannot access external code repositories such as GitHub, you can use the mirroring method:

  1. First, download the files to your local machine.

  2. Upload the code to a storage space you can access (for example, upload to OSS).

  3. Obtain the new download link and send it to OpenClaw to install, following the steps in Install from a URL.

Does OpenClaw support loading Skills from custom folders? What is the priority order for loading Skills?

Yes. You can add extra directories using ~/.openclaw/openclaw.json and the skills.load.extraDirs field (lowest priority).

The default Skill loading priority is: <workspace>/skills > ~/.openclaw/skills > built-in > skills.load.extraDirs.

The clawhub tool installs Skills to ./skills by default. OpenClaw treats this as <workspace>/skills.

How do I restart the OpenClaw Gateway service?

Connect remotely to your Simple Application Server and run the Gateway helper command in the terminal:

openclaw gateway restart

What are the different configuration files in OpenClaw installed on a simple application server used for?

• Configuration file directory: /root/.openclaw/

• Main configuration file: /root/.openclaw/openclaw.json (contains the models in use)

• Model API key configuration file: /root/.openclaw/agents/main/agent/auth-profiles.json

You can directly edit these configuration files to add models or configure API keys.

How do you manually configure and use a Coding Plan purchased from Model Studio Coding Plan in OpenClaw?

1. Copy and save the Coding Plan’s plan-specific API key.

2. In the following code block, replace "apiKey": "YOUR_API_KEY" with your plan-specific API key.

    "models": {
       "mode": "merge",
       "providers": {
         "bailian": {
           "baseUrl": "https://coding-intl.dashscope.aliyuncs.com/v1",
           "apiKey": "<span class="var-span" contenteditable="true" data-var="YOUR_API_KEY">YOUR_API_KEY</span>",
           "api": "openai-completions",
           "models": [
             {
               "id": "qwen3-max-2026-01-23",
               "name": "qwen3-max-thinking",
               "reasoning": <span class="hljs-keyword">false</span>,
               "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
         }
       }
     },

3. On the Simple Application Server console – Servers page, click the instance ID on the card of the server where OpenClaw is deployed to go to the Server Overview page.

4. Click the Application Details tab. On the Access Control page, click Execute Command on the right side of the Open Website page. Then, click the website URL link in the pop-up window to open the OpenClaw conversation page.

5. Click Config > All Settings > Raw to open the configuration file. Copy the code block with the modified "apiKey" parameter. Replace the original "agents"{...} content in your configuration code as shown in the following image, and then save the changes.

How do I configure Brave Search for web search in OpenClaw?

Web search using the built-in SearXNG Skill is enabled by default. You can configure the Brave Search API for web search in Hong Kong (China) and regions outside China. For more information, see the Brave Search documentation or the Alibaba Cloud document Deploy the OpenClaw runtime image and build a DingTalk AI employee.

If you already have a Brave Search API key, log on to your Simple Application Server and run the following command to configure Brave Search:

openclaw configure --section web

How do I resolve the "command not found: openclaw" error in the Simple Application Server terminal?

Create a symbolic link by running the following commands:

ln -sf /home/clawdbot/dist/entry.js /usr/bin/openclaw
openclaw --help

Does the configurable Brave Search API support web search?

You can configure Brave Search API for web search in Hong Kong (China) and regions outside China. This feature is not supported on servers in the Chinese mainland.

1. Create a Brave Search API account and generate an API key on the Brave Search official website.

2. Configure the image. In the OpenClaw page, click Config > All Settings > Raw in the navigation pane on the left to open the configuration file. Replace BRAVE_API_KEY with your Brave Search API key, and copy the following code block into the configuration file.

   tools: {
       web: {
         search: {
           provider: "brave",
           apiKey: "BRAVE_API_KEY",
           maxResults: 5,
           timeoutSeconds: 30,
         },
       },
     },

How do I use Docker containers to run tools in OpenClaw?

OpenClaw supports running tools in Docker containers to mitigate potential risks. This is an optional feature that, when enabled, runs tools in an isolated sandbox environment. For more information, see the OpenClaw sandboxing feature guide.

What are the common command-line interface (CLI) tools in OpenClaw?

Log on to the server terminal and switch to the root user to run the following commands.

How do I configure OpenClaw to use only the free quota for Model Studio in the Singapore region and avoid extra charges?

By default, Model Studio Large Language Models (Singapore region) incur charges once the free quota is exhausted. To avoid extra model invocation charges, enable Stop when free quota is exhausted in the Model Studio console. After enabling this setting, the model will become unavailable once the free quota is used up. You can then switch to another model with an available free quota.

Are there server configuration requirements for purchasing OpenClaw? Can I select the OpenClaw image on any configuration?

You should select a configuration with at least 2 vCPUs and 2 GB of memory to ensure optimal service performance.

Will there be other fees after purchasing a Simple Application Server instance with the OpenClaw application image deployed?

If you configure OpenClaw on an instance to invoke models using an API key from Model Studio, you will be charged based on token usage. For billing details, see the Model Invocation Billing documentation on the Model Studio platform.

How can I check the remaining free quota for models in Model Studio for the Singapore region?

Log on to the Model Studio console. In the Free Quota area, you can view the remaining free quota for the model under your account.

Why am I still being charged for model API calls after subscribing to Coding Plan?

Possible reasons are as follows:

1. If you configured your AI tool with the general-purpose Model Studio API key (in the format sk-xxxxx) and the general-purpose base URL (which does not contain the keyword "coding"), the system identifies these calls as pay-as-you-go requests for Model Studio. These calls cannot be deducted from your Coding Plan quota and will generate pay-as-you-go bills. You must correctly configure the plan-specific credentials. For more information, see 2. Get the plan-specific API key and base URL.

2. The charges may have been incurred from model API calls that you made before you subscribed to Coding Plan. Because bills are settled hourly, there can be an hour-level delay during peak periods. For example, if you call the API at 16:00 and then subscribe to and use Coding Plan at 19:00, the bill generated at 19:30 will show the charges for the call made at 16:00. The actual billing time is the one that applies. For more information, see Bill query and cost management.

How to view model invocation records in Alibaba Cloud Model Studio?

What are the differences between Alibaba Cloud Model Studio's International (Singapore), US (Virginia), and the Chinese mainland (Beijing) regions?

Alibaba Cloud Model Studio offers model services in the International (Singapore), US (Virginia), and Chinese mainland (Beijing) regions. Choosing a nearby region reduces network latency. Each region has a unique service endpoint (Endpoint/Base URL), and API keys are not interchangeable. Supported models, platform features, and pricing also differ. For details, see the Model List.

What should I do if the OpenClaw Chat page returns no content or does not respond?

1. Check whether the API key is configured correctly: Connect remotely to your Simple Application Server. In the code block below, replace the API key and Base URL with the values for your region. Copy and paste the updated code into the terminal to test the model call. If an error occurs, search for it in the error message documentation and apply the recommended solution. Base URLs are not interchangeable across regions.

   • China (Beijing): https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions

   • US (Virginia): https://dashscope-us.aliyuncs.com/compatible-mode/v1/chat/completions

   • Singapore: https://dashscope-intl.aliyuncs.com/compatible-mode/v1/chat/completions

   • Coding Plan: https://coding-intl.dashscope.aliyuncs.com/v1

   curl -X POST YOUR_API_KEY_BASE_URL \
   -H "Authorization: Bearer YOUR_API_KEY" \
   -H "Content-Type: application/json" \
   -d '{
       "model": "qwen3-max-2026-01-23",
       "messages": [
           {
               "role": "user",
               "content": "Who are you?"
           }
       ]
   }'

2. Check whether the model service is overdue or rate-limited.

   1. Coding Plan: Each plan has request limits per 5 hours, per week, and per month. You can check your quota usage in the Coding Plan console. When you exceed the limit, the error hour/week/month allocated quota exceeded appears.

      Solution: Wait for the quota to reset automatically or upgrade to the Pro plan.

   2. Pay-as-you-go Model Studio API key:

      1. Check your free quota. Log on to the Model Studio console. In the Free Quota area, you can view the remaining free quota for the model under your account.

      2. Check for overdue payments. Go to the Expenses and Costs center to ensure your account has no overdue payments.

3. Review error logs: Go to the OpenClaw Chat page. In the left navigation pane, click Logs. Select the WARN and ERROR boxes to view the error logs. Search for the error in the OpenClaw FAQ document to find a solution.

   

   If no errors appear in the Logs, connect remotely to your Simple Application Server and run the following command in the terminal to restart the Gateway service.

   moltbot gateway restart

4. Reset the image.

   If you confirm that your server contains no important data, consider resetting the Simple Application Server image to the latest OpenClaw application image and reconfiguring it.

   Important

   Resetting the operating system is equivalent to reinstalling the operating system. This deletes all data on the system disk, including saved configurations, logs, and databases. Before proceeding, back up important data. We recommend creating a snapshot or exporting data to a local device.

How do I resolve the "HTTP 401: invalid access token or token expired" error on the OpenClaw page?

The error indicates that your token has expired. Go to the Application Details page in the Simple Application Server console, regenerate the OpenClaw login URL, and then try accessing it again.

Why does OpenClaw show a "blocked" message when installing Skills?

The OpenClaw management interface supports one-click installation for some Skills. However, most Skills require Homebrew, so one-click installation is not supported. To use these Skills, you must install Homebrew on your server and then debug manually.

How do I resolve the "disconnected (1008): unauthorized" error when accessing the OpenClaw website?

This occurs because your access link is missing an authentication token. The OpenClaw web console does not allow direct access using an IP address, and the URL must include the correct token parameter. In the server console, click the instance ID on the server card to go to the server overview page. Click the Application Details tab, and in the Access Control page, click Execute Command to obtain the correct access URL that includes the token.

How do I resolve the "disconnected (1008): control ui requires HTTPS or localhost (secure context)" error when I access the OpenClaw website?

This error occurs because the token is not initialized. Go to the Application Details page in the Simple Application Server console. Set the API key as instructed in step 2, and then obtain a new access link to visit the website.

How do I resolve the "disconnected (1006): no reason" error when accessing the OpenClaw website?

Follow these steps to troubleshoot:

1. Regenerate the token. Log on to the Alibaba Cloud Simple Application Server console. Find your instance (ID: t-gz06aj39x9e5ukg), and go to the Application Details page to generate a new token.

2. Append the new token to the endpoint in the format http://XXX:18789/?token=your_new_token, and open the URL in an incognito window for testing.

3. Confirm service status. Use SSH to log on to the server, and verify that the OpenClaw service is running and listening on port 18789.

4. Check security group rules. Ensure that the security group of your Simple Application Server allows inbound TCP traffic on port 18789.

   If the service is not properly bound to a public IP or if reverse proxy settings are incorrect, WebSocket connections may fail. Test connectivity to 8.138.139.147:18789 first using curl or telnet.

How do I resolve the "Failed to discover Alibaba Cloud models: 401 Unauthorized" error?

This issue may occur because of an overdue payment or insufficient model invocation permissions. To test Model Studio model calls and review the detailed error to confirm whether the API key works, see the Check API Key step.

How do I resolve the "command not found: openclaw" error in the Simple Application Server terminal?

Create a symbolic link by running the following commands:

ln -sf /home/clawdbot/dist/entry.js /usr/bin/openclaw
openclaw --help

What should I do if I cannot get the token after I enable Tailscale?

This happens because Tailscale modifies routing policies. You can try using the following command to read the token information directly from within the instance:

1. Log on to the OpenClaw server terminal.

2. Run the following command:

   echo $(sed -z 's/.*"token": "\([^"]*\)".*/\1/' /root/.clawdbot/clawdbot.json | tr -d '\0')

3. The string that is output in the terminal is your token.