ZCode

Prerequisites

• You have signed up for a Huawei Cloud account. For details, see Signing Up for a Huawei Cloud Account.
• You have subscribed to a built-in service or created an endpoint. For details, see Subscribing to a Built-in Service and Creating an Endpoint.

Billing

When using the MaaS model service, you are charged based on the number of tokens used. For details, see MaaS Text Generation Models.

Constraints

This function is only supported in the CN-Hong Kong region.

Step 1: Install ZCode

Download the ZCode version for your operating system from the ZCode official website and install it. ZCode supports macOS, Windows, and Linux.

After the installation is complete, open ZCode and choose to connect to a Z.ai account or a BigModel account, or use an API key to access the model. This section provides guidance on using an API key to access a MaaS model.

Step 2: Access MaaS Model APIs

1. Open ZCode and choose Choose model > Manage models.
   Figure 1 Model settings
   

2. On the Model settings page, click Add provider in the upper right corner and select Create custom provider.
   Figure 2 Add a provider
   

3. On the New provider page, set the parameters.
   Figure 3 Add a custom provider
   

   Table 1 Custom provider settings

   Parameter	Description	Example
   Provider name	Name of the custom provider. Click in the upper right corner and choose Rename.	Huawei Cloud MaaS
   Base URL	(Mandatory) Enter a base URL. For the Chat Completions format, enter the following URL: https://api-ap-southeast-1.modelarts-maas.com/v2 https://api-ap-southeast-1.modelarts-maas.com/openai/v1 For the Anthropic Messages format, enter the following URL: https://api-ap-southeast-1.modelarts-maas.com/anthropic	https://api-ap-southeast-1.modelarts-maas.com/openai/v1
   API format	(Mandatory) Select either of the following API formats: Chat Completions Anthropic Messages	Chat Completions
   API Key	(Mandatory) Enter an API key. You can use an existing API key. If no API key is available, obtain an API key.	gA4ocs*****Tc-Kz4AQ

4. Click Add model and configure the model parameters.
   Figure 4 Model configuration example
   

   Table 2 Model settings

   Parameter	Description	Example
   Model ID	(Mandatory) Enter the model ID. Obtain the value from Model Parameter in the model list.	deepseek-v4.1-flash
   Context window	(Mandatory) Specify the context length that can be processed by a model at a time, in tokens. The value can be configured to the maximum input length supported by the model. You can obtain the Max input length from the model list. Do not exceed the input length limit of the model.	1,000,000
   Maximum output tokens	(Mandatory) Enter the maximum number of tokens that can be generated for a single request, in tokens. The value can be configured to the maximum output length supported by the model. You can obtain the Max output length from the model list. Do not exceed the output length limit of the model.	384,000
   Advanced settings	(Optional) You can set the Input types, Model capabilities, and Reasoning levels.	-

5. Click Save and then click Test model in the operation column. If the connection is successful, the model was added successfully.
   Figure 5 Testing model connectivity
   

   

Step 3: Use ZCode

1. In the navigation pane of the ZCode home page, choose New task. Then select the model configured in Step 2: Access MaaS Model APIs.
   Figure 6 Selecting a model
   

2. Start a dialogue in the dialog box to experience the model.
   Figure 7 Example conversation
   

Common Errors