Skip to main content

Quickstart

Create and deploy your first High-Code Application from the console: create the application, complete cloud authorization, configure deployment, and verify with tests.

This tutorial walks you through creating and deploying a High-Code Application from the console in about 10 minutes.

Step 1: Create the Application

  1. Go to the console's High-Code page (left navigation: Sandbox > High-Code) and click Create Application.
  2. In the Create High-Code Application dialog:
    • Type: choose Console creation (upload code packages quickly through a visual interface—suitable for prototype validation; CLI creation suits automated production, see Develop & Deploy).
    • Application Name: required; this tutorial uses "My First High-Code Application" (up to 50 characters).
    • Description and Avatar: optional.
  3. Click Create Now to enter the deployment configuration page.
During creation, follow the page guide to authorize the related roles for Function Compute (FC) and API Gateway. If permissions are insufficient, contact your account or IT administrator to grant the required authorization.

Step 2: Configure the Deployment

On the deployment configuration page, complete the following: Select deployment mode
  • Serverless Function (default): for low-load, stateless, fast-startup scenarios with low deployment cost.
  • K8s: for high-performance, stateful Agents running long tasks. Requires the ACK container service and authorization.
Submit code
  • Use template code (default): pick a preset application template to get started quickly.
TemplateDescription
Basic chat AgentA basic LLM-chat example Agent, ideal for quick trials and secondary development.
Tool-calling AgentBuilt-in MCP tools (web search, calculator, etc.) demonstrating tool-calling capabilities.
Deep research AgentA deep research Agent built with AgentScope components; conducts internet research and generates reports.
  • Upload code package: upload a locally developed .whl package; see Develop & Deploy for packaging.
Select deployment resources Configure the specification (vCPU, memory, disk), minimum instances, per-instance concurrency, and deployment region as needed. Defaults are fine for a first trial.
For latency-sensitive workloads, set the minimum instance count to ≥ 1 for millisecond-level warm starts and uninterrupted service. See Instance types and specifications.
Enable observability (optional): once enabled, you can analyze the call paths of your application code and monitor call parameters.

Step 3: Deploy and Wait for the Build

After configuration, click Deploy Now to start the deployment. The build and deployment typically take a few minutes. The page shows real-time progress and elapsed time; click the Logs link next to the progress bar for detailed build and deployment logs.

Step 4: Test the Application

After deployment completes, configure your Bailian API Key, then use the test panel on the right. See the API Key management page for how to obtain a key.
  • Text chat experience: interact with the application directly to verify functionality.
  • API testing: manually call the application's API endpoints (such as GET /health, /process)—suitable for development debugging.

Next Steps