Frequently asked questions about deploying OpenClaw (Clawdbot) on Elastic Desktop Service (EDS).
Is the one-click OpenClaw (Clawdbot) deployment on EDS the same as native Clawdbot?
The one-click deployment is a repackaged version of native Clawdbot. It integrates the core Clawdbot logic with a dedicated EDS image, providing a pre-configured runtime that combines a native kernel with an automated shell. Use the built-in configuration tool to get started. No additional deployment or installation is needed.
Can I deploy OpenClaw (Clawdbot) on an existing EDS Personal Edition instance?
Yes. Change the image to use OpenClaw (Clawdbot). See One-click deployment of OpenClaw (Clawdbot) on EDS.
I already have cloud computer core-hours and don't want to use the OpenClaw image. Can I set up my own environment?
Yes. Deploy Clawdbot natively on a cloud server.
I receive duplicate messages from DingTalk.
Upgrade to the latest image version. See Deploy OpenClaw (Clawdbot) on Elastic Desktop Service.
Error: "Stream mode connection failed. Please refer to the Stream Mode SDK Access Guide."
This error usually occurs because the APP KEY or APP SECRET for the DingTalk channel is incorrect. Check your configuration.
Error: HTTP 401: Incorrect API key provided.
Log on to your EDS instance and run clawdbot gateway restart in the terminal to restart the service.
After the service starts, disable the Lark channel, then re-enable it and test.
How do I switch models?
During model configuration, click the Switch to general-purpose model button in the upper-right corner.
The command curl ifconfig.me fails to get the IP address.
This command gets the public IP address of the EDS instance. Try curl cip.cc instead.
QQ replies with the message: "The robot has gone to Mars. Please try again later."
Possible reasons:
The IP address whitelist is not added on the QQ robot configuration page.
The App ID and App Secret for the QQ robot are not added to OpenClaw.
The wrong App ID and App Secret for the QQ robot were added.
For configuration details, see QQ integration.
QQ does not reply to messages.
Check the following items and perform the required operations:
Check the logs for any error messages. If you find a similar message, check whether the Application ID (APPID) and BOT SECRET in the QQ Channel configuration are correct.
If the configuration is correct or no related logs are found, regenerate the AppSecret (BOT SECRET). Replace the existing BOT SECRET with the new one, click Save Configuration, and retry.
How do I check if my configuration file is correct or has errors?
Run clawdbot doctor. Any issues are reported as error messages.
After connecting to DingTalk, Lark, or QQ, can I send images or files in the chat for OpenClaw to process?
No. OpenClaw cannot read images or files sent through chat.
How do I use a Coding Plan key with EDS Personal Edition?
To use a Coding Plan API key, click Switch to general-purpose model. Set Model Name to qwen3-coder-plus or qwen3-max-2026-01-23. See the Coding Plan documentation. Change the BASE URL to https://coding.dashscope.aliyuncs.com/v1. Click Test Connection. The connection succeeds if the test returns a normal status.
If you cannot find the Switch to general-purpose model option, update the image to the latest version.
Can I deploy OpenClaw on EDS Enterprise Edition?
Yes. See Deploy OpenClaw (Clawdbot) on Elastic Desktop Service.
Can I upgrade only the OpenClaw version on EDS Personal Edition?
Upgrading only the OpenClaw version is not supported. The EDS Personal Edition image is a fully tested and integrated package. Do not use commands such as clawdbot update to upgrade, as this may cause environment errors.
An error occurs when using sk-sp-*** in the Model Studio LLM configuration.
Problem
Error message: "Authentication failed: {"error":{"code":"invalid_parameter_error","message":"model `qwen-max-latest` is not supported.","param":null,"type":"invalid_request_error"},"request_id":"a3***5"}"
Solution
To use a Coding Plan API key, click Switch to general-purpose model. Set Model Name to qwen3-coder-plus or qwen3-max-2026-01-23. See the Coding Plan documentation. Change the BASE URL to https://coding.dashscope.aliyuncs.com/v1. Click Test Connection. The connection succeeds if the test returns a normal status.
If you cannot find the Switch to general-purpose model option, update the image to the latest version.