New to Qwen Cloud? Get up and running in under 30 minutes with these resources. Whether you're making your first API call or choosing the right model for your project, these guides will fast-track your understanding of the platform.

Quick Start in 3 Steps

  1. Register on Devpost (2 minutes)
  2. Sign up for Qwen Cloud — Get your free trial and request hackathon credits via the coupon form. (5 minutes)
  3. Join the Qwen Cloud Discord (2 minutes)

Pick your track, review the sample projects and reference architecture below, and start building. You can have your first agent running in about 30 minutes.

Essential Documentation

Resource

URL

Description

Introduction to Qwen Cloud

https://bit.ly/intro-qwencloud

Platform overview, capabilities, and what you can build.

First API Call

https://bit.ly/qwencloud-first-api

Set up your API key and make your first request in Python, Node.js, or curl.

Choose Your Model

https://bit.ly/qwencloud-modelselection

Pick the right Qwen model for your track and use case.

Pricing & Free Credits

https://bit.ly/qwencloud-pricing

Understand token pricing and how to use your hackathon credits.

Get an API Key

https://bit.ly/qwencloud-getapi

Generate and manage your API credentials securely.

API Base URL: https://dashscope-intl.aliyuncs.com/compatible-mode/v1

The API is OpenAI-compatible, so you can use the OpenAI SDK in Python or Node.js.

Inspiration

Not sure where to start? Here are jumping-off points and project ideas for each of the five hackathon tracks.

Track 1: MemoryAgent

Build agents with persistent, queryable memory that retain and recall information across sessions.

Project Ideas:

  • A research assistant that remembers every paper it has read and can synthesize findings across sessions.
  • A customer support agent that recalls full conversation history and user preferences without re-asking.
  • A personal knowledge base that indexes everything you share and surfaces relevant context proactively.

Track 2: AI Showrunner

Create agents that produce, direct, or orchestrate creative content at scale.

Project Ideas:

  • An automated podcast pipeline that researches topics, writes scripts, and coordinates multi-voice narration.
  • A narrative game engine where AI agents dynamically generate storylines based on player choices.
  • A multi-agent writers' room that brainstorms, drafts, critiques, and revises content collaboratively.

Track 3: Agent Society

Design systems where multiple agents collaborate, negotiate, or compete to solve problems.

Project Ideas:

  • A simulated marketplace where buyer and seller agents negotiate prices and terms autonomously.
  • A multi-agent debate platform where agents argue different positions and a judge evaluates arguments.
  • Cooperative problem-solving swarms that divide complex tasks and coordinate to reach solutions.

Track 4: Autopilot Agent

Build agents that autonomously complete complex, multi-step workflows from start to finish.

Project Ideas:

  • A code review pipeline that analyzes PRs, identifies issues, suggests fixes, and verifies corrections.
  • An autonomous data analyst that ingests datasets, identifies patterns, generates visualizations, and writes reports.
  • A self-healing deployment system that monitors services, diagnoses failures, and applies fixes automatically.

Track 5: EdgeAgent

Deploy lightweight, efficient agents optimized for constrained environments and real-time performance.

Project Ideas:

  • An on-device personal assistant that works without constant cloud connectivity.
  • An offline-capable field agent for remote workers who need AI assistance in low-bandwidth environments.
  • A latency-sensitive real-time application that processes sensor data and makes instant decisions at the edge.
Technical Resources