Use Cline VSCode Plugin with Qwen for AI Code Assistance - Model Studio

Cline is an intelligent programming plugin for VSCode. Integrate models from Alibaba Cloud Model Studio to complete complex programming tasks.

Prerequisites

Get your API key and API host. Ensure that you have activated the model service in Alibaba Cloud Model Studio.
In the Model list, select the Qwen model that you want to use.
Because Cline's features are subject to change, the list of supported models may vary.
Download and install VSCode.

Steps

1. Install Cline

Open VSCode. In the extension marketplace, search for Cline and install it.

2. Configure models and API keys

After installation, click the icon in the sidebar to open the Cline interface. Click Bring my own API key. In the dialog box that appears, select an API Provider:

Recommended: Select OpenAI Compatible to use a wider range of models, including the latest ones.
Alternative: Select Alibaba Qwen. The configuration is simpler, but this option may not support some of the latest models.

If you have used Cline before, you can change the configuration by clicking the Settings button in the upper-right corner.

Important

This document applies only to pay-as-you-go mode. Coding Plan users must use your exclusive base URL and API key instead. For details, see Coding Plan for Cline.

OpenAI Compatible

This method is recommended because it supports a wider range of models. Configure the settings as follows:

Configuration item	Description
API Provider	Select OpenAI Compatible.
Base URL	Enter the URL for the region where you want to invoke the model: Singapore region: `https://dashscope-intl.aliyuncs.com/compatible-mode/v1` Virginia region: `https://dashscope-us.aliyuncs.com/compatible-mode/v1` Beijing region: `https://dashscope.aliyuncs.com/compatible-mode/v1`
API Key	Enter the API key for the region where you use the model.
Model ID	Enter the name of the model you want to use. We recommend that you select a model with strong programming capabilities, such as: `qwen3.5-plus` `qwen3-coder-next` `qwen3-coder-plus` `qwen3-max` For models supported by each pattern, see Model List.
Others	Keep other options as default. Explore more features as needed. Note If you use Qwen3 (thinking mode) or QwQ models, click `MODEL CONFIGURATION` and check Enable R1 messages format.

Click Continue to complete the configuration.

If you have used Cline before, click Done in the upper-right corner.

Alibaba Qwen

Configuration Item	Description
API Provider	Select Alibaba Qwen.
Alibaba API Line	If you use a Singapore region model, select International API. If you use a Beijing region model, select China API.
Qwen API Key	Enter the API key for the region where your model is deployed.
Model ID	From the drop-down list, select the model that you want to use. We recommend selecting a model with strong programming capabilities, such as: `qwen3-coder-plus` `qwen3-coder-480b-a35b-instruct` Some of the latest models (such as the qwen3-coder series) may not be selectable in the Beijing region. If you want to use these models, we recommend that you change the API Provider to OpenAI Compatible.
Other	Keep the default settings for the other options. You can explore additional features as needed. Note For more information about model context parameters and prices, see Model list.

Click Continue to complete the configuration.

If you have used Cline before, click Done in the upper-right corner.

20251201174144

3. Use Cline

You can use the Auto-approve option above the input box to grant Cline permission to read, write, and execute commands.

Important

Enabling these permissions increases token consumption. Before you enable them, ensure that you understand Cline's features.

Cline operates in two modes: Plan and Act.

In Plan mode, Cline primarily collects information, breaks down problems, and designs a detailed plan to complete the task. It does not directly modify files.
In Act mode, Cline focuses on creating a solution and then performs the required operations. It can run commands and modify files.

You can switch between modes at the bottom of the dialog box.

Plan

For example, to generate a Python quicksort algorithm, you can ask Cline in Plan mode: "Write a Python quicksort algorithm and provide the code directly."

20251127145130

In Plan mode, Cline does not create or modify code files. In Act mode, Cline creates a new file and writes the code into it.

Act

Start a new session and switch to Act mode. For example, to generate a Python quicksort algorithm, you can ask Cline: "Write a Python quicksort algorithm."

20251127151539

Cline generates the code and writes it to a file. It then asks whether you want to run the file for testing. Cline runs the command only after you provide confirmation.

FAQ

Q: Why do I get a 401 Incorrect API key provided error?

A: This error can occur for the following reasons:

The region of your API key does not match the region of the Base URL. Ensure that they are in the same region.
An older version of Cline does not support selecting Alibaba Qwen as the API Provider. To resolve this, upgrade Cline or change the API Provider to OpenAI Compatible.

Q: Why is my token consumption high?

A: Cline reads your local project files and makes multiple API calls, which consumes many tokens. For simple question-and-answer scenarios, clear the Auto-approve checkbox and use Plan mode.

Q: How can I reduce token usage?

Reduce irrelevant files: To avoid scanning unrelated files and consuming tokens unnecessarily, start Cline from your specific project folder, which should contain only essential project files.
Summarize conversations: Cline uses your conversation history as context. Click the Compact Task button or enter the command /compact to summarize the conversation.
Use precise instructions: Vague requests can trigger unnecessary file scans and consume more tokens. Provide clear, specific questions or exact instructions when you use Cline.
Break down tasks: For complex tasks, split them into several simple jobs.

Q: Why do I get this error when using the QwQ model: 400 <400> InternalError.Algo.InvalidParameter: An error during model pre-process?

A: In the Settings interface, click MODEL CONFIGURATION and select the Enable R1 messages format checkbox.