All Products
Search

This Product

    Platform For AI:Deploy your dedicated QwenPaw on PAI-EAS in 5 minutes

    Document Center

    Platform For AI:Deploy your dedicated QwenPaw on PAI-EAS in 5 minutes

    Last Updated:Jun 22, 2026

    Deploy QwenPaw on PAI-EAS to quickly build a dedicated AI assistant powered by custom PAI-EAS models or Model Studio models. QwenPaw is an open-source, local-first, self-hostable agent framework released by the AgentScope team in February 2024 that supports inference cost control through a Coding Plan.

    Deploy the EAS service

    1. Log on to the PAI console. Select a region on the top of the page. Then, select the desired workspace and click Elastic Algorithm Service (EAS).

    2. On the Inference Service tab, click Deploy Service. In the Scenario-based Model Deployment area, click QwenPaw Deployment.

    3. Configure the key parameters:

      Parameter

      Description

      Service Name

      Example: qwenpaw_demo.

      Image Configuration

      Select the latest version.

      Resource Configuration

      Select a CPU instance type based on your resource requirements. The default is ecs.c7a.large.

      Model Settings

      Select an OSS path to persist the QwenPaw configuration, such as oss://examplebucket/qwenpaw/.

      VPC

      Create and configure a VPC, vSwitch, and security group.

      Note

      QwenPaw requires public internet access for search and browser functions. We recommend selecting a VPC that is already configured for public network access.

    4. Click Deploy and wait for the service status to change to Running. This indicates that the deployment is successful.

    Configure public network access

    EAS services access the public internet through an Internet NAT Gateway. For details, see Internet NAT Gateway.

    1. Create a NAT gateway and bind an EIP

    Go to the NAT Gateway - Internet NAT Gateway purchase page.

    • Billing Method: Select pay-as-you-go.

    • Region: Select the region where your EAS service is located.

    • Network and Zone: Select the VPC configured for your EAS service. This setting cannot be changed after creation.

    • EIP: Select an EIP that is not associated with an instance, or purchase a new one.

    2. Configure an SNAT entry

    Go to the Internet NAT Gateway page. Find the Internet NAT Gateway that you created. In the Actions column, click Configure SNAT , and then click Create SNAT Entry.

    • SNAT Entry: Select VPC granularity.

    • Select EIP: Select the configured EIP.

    Start the web UI

    1. Click the image button in the Invoke/Log/Monitoring column to open the web UI page.

    2. Configure models. On the Settings > Models page, configure a Provider. In the LLM Configuration section, select a provider and model.

      Aliyun Coding Plan

      Go to the (international) to subscribe to the service and get your dedicated API key.

      On the model settings page, find the Aliyun Coding Plan provider card and click Settings in the lower-right corner. In the dialog box, the Base URL is pre-filled as https://coding.dashscope.aliyuncs.com/v1. Paste your API key into the API key field and click Save.

      EAS model service

      1. Click Add Provider. For Protocol, select OpenAI-compatible.

      2. Set the URL and API key. See Get the endpoint and token and append /v1 to the endpoint. Enter the full URL in the Base URL field and the service token in the API key field. Click Test Connection to verify, and then click Save.

      3. Add a model. On the PAI-EAS provider card, click Models, and then click + Add Model.

      DashScope

      On the DashScope provider card, click Settings and add your Model Studio API key. See Select a model to add more models.

      In the Configure DashScope dialog box, the Base URL is pre-filled with the default value. Paste your API key into the API key field, click Test Connection to verify, and then click Save.

    3. Start a conversation with or assign a task to QwenPaw.

      For example, select the Qwen3.5 Plus model and enter "weather in Hangzhou tomorrow" in the chat box. QwenPaw performs multi-step thinking and uses the browser_use tool to automatically retrieve web information and return the weather forecast, including conditions, temperature range, and wind level.

    Channel configuration

    QwenPaw supports integration with channels such as DingTalk and Lark. For details, see Channel Configuration.