Qwen
Officially listedAlibaba's Qwen large-model AI assistant
Qwen
Qwen (千问) is Alibaba's flagship AI assistant built on the Qwen family of large language models, positioned as a personal AI assistant that both chats and gets things done. In November 2025 it moved to its own domain, qianwen.com, extending the mobile experience to the web.
Main features
- Multimodal conversation: text Q&A, image understanding and generation, video understanding, and voice interaction
- Content creation: stories, formal documents, emails, scripts, and poetry
- Productivity: slide generation, document translation, long-text analysis, information extraction, and summarization
- Coding: powered by Qwen3-Coder for code generation and optimization
- Real-time meeting notes with smart cleanup
Qwen performs strongly in public evaluations — Qwen3.0 surpasses GPT-4 in comprehensive benchmarks, beats Grok3 in code generation, and clearly exceeds Gemini 2.5 Pro and OpenAI o1 in agent capability tests. It also integrates with Amap for trip planning. The consumer app and website remain free with no usage caps on core features; developers pay per token via the API.
Pricing
## 价格方案 ### 个人使用版（完全免费） - 千问APP和网页版完全免费使用 - 无使用次数限制 - 所有核心功能开放：智能问答、文档创作、图像生成、视频生成、代码编写等 ### API调用版（开发者按量付费） - **免费额度**：新用户首次开通可免费领取千万tokens，每个主流模型限免100万tokens - **qwen-turbo**：￥2/百万tokens（输入），￥6/百万tokens（输出） - **qwen-plus**：￥4/百万tokens（输入），￥12/百万tokens（输出） - **qwen-max**：￥40/百万tokens（输入），￥120/百万tokens（输出） **注**：普通用户使用千问APP或网页版（qianwen.com）完全免费，仅开发者通过API调用模型时需按Token付费。
FAQ
Is Qwen free to use?
Yes. The Qwen app and qianwen.com are completely free with no usage limits on core features. Only API usage by developers is billed per token.
How strong is the Qwen model?
Qwen3.0 surpasses GPT-4 in comprehensive benchmarks, beats Grok3 in code generation, and significantly exceeds Gemini 2.5 Pro and OpenAI o1 in agent capability tests. Qwen-derived models on Hugging Face exceed 170,000, the largest open-model family.
Can Qwen write code?
Yes. Coding is powered by the integrated Qwen3-Coder model for code generation, optimization, and programming assistance.