Qwen products: one model family, every surface

One foundation, eight products

Everything in the Qwen ecosystem traces back to a single family of models - what changes is the surface you meet them on.

Qwen is the family of large language models built by Tongyi Lab, the AI division of Alibaba Cloud, first released in 2023 under the Chinese name Tongyi Qianwen. Over the years it has grown from a single chat model into one of the most widely deployed model families in the world, and Alibaba has wrapped those models in a suite of products aimed at very different audiences - casual users who just want to ask questions, developers shipping production software, researchers digging through literature, and creators generating video and voice. The unifying idea is simple: build the strongest possible foundation models, then meet people wherever their work actually happens.

That is why there is no single "Qwen app" you must learn. Instead there are eight core products, each tuned to a job. A few you open in a browser and never think about the plumbing; others you install, script against, or wire into a production pipeline. The same account links them all, so a custom prompt you save in the chat, a voice you clone, or an API key you generate is available everywhere you sign in. Below is the map, followed by a closer look at each one.

Qwen Chat

The front door for everyone - a free, no-install AI workspace that does far more than answer questions.

Who it's for

Students summarizing papers, writers drafting and editing across 130+ languages, analysts pasting a CSV for instant insight, developers debugging a stack trace, anyone who wants a capable assistant without touching a line of configuration. The free tier is generous; an optional Pro tier removes daily limits, raises upload sizes, and adds priority access. Because it is free to start and requires no account to try, Qwen Chat is also the cleanest way to evaluate whether Qwen's quality fits your needs before you commit to anything heavier.

Qwen Code

A coding assistant that understands whole repositories, not just the file in front of it.

Getting started

Install the VS Code extension or JetBrains plugin from the marketplace, or grab the command-line tool through your package manager. The CLI is the natural choice for terminal-centric workflows and agentic, multi-step coding tasks where you want the model to run an edit-test-fix loop rather than autocomplete a line. For teams that want full control, the open GitHub repository hosts the tooling so you can inspect, fork, and integrate it into your own pipelines.

Qwen Research

A deep-research toolkit for the moment a single answer isn't enough and you need a sourced report.

Access is through a dedicated web platform and via API for teams that want to embed research workflows into their own products. It pairs naturally with the broad knowledge and long-context strengths of the underlying models: a large context window means the agent can hold many documents in mind at once instead of summarizing each away and losing detail.

The API Platform

The developer's entry point: production-grade access to the entire Qwen catalogue through one OpenAI-compatible endpoint.

A first request

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_DASHSCOPE_API_KEY",
    base_url="https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
)

completion = client.chat.completions.create(
    model="qwen-plus",
    messages=[{"role": "user", "content": "Hello, Qwen!"}]
)
print(completion.choices[0].message.content)

The model catalogue at a glance

All of these are reachable through the same key and endpoint - you change only the model string. Pricing varies by region and changes often, so always confirm current rates in Model Studio before committing a production workload.

Creative media: Video & Voice

The same foundation, pointed at pixels and sound - turning prompts into cinematic clips and natural speech.

Mobile app & Android APK

Every Qwen capability in your pocket - with a direct-download option for regions and devices without an app store.

One account, every Qwen tool

The ecosystem is designed so the products feel like one product wearing different clothes.

A single Qwen account works seamlessly across Chat, Code, Voice, the Video Generator, the mobile app, and the API Platform. Sign in once and your chats, files, and prompts sync across web, desktop, and mobile; you manage API keys, usage, and billing from one dashboard; and custom voices, video styles, and system prompts you save are available wherever you go. New accounts come with free API credits to get started, and the platform supports two-factor authentication and enterprise SSO for organizations that need it.

That shared backbone is the quiet advantage of the Qwen approach. Because every product is built on the same foundation models and the same account system, the skills and assets you develop in one carry over to the rest. A prompt pattern that works well in the chat translates to the API; a voice you clone is usable in the video generator; the model you trust in one place behaves consistently across the others. For individuals that means less friction; for teams it means a single vendor relationship, one security posture, and one billing relationship across a surprisingly broad set of capabilities.

Frequently asked questions