Qwen Team (Alibaba)
To provide an AI assistant for everyone that enables creativity, collaboration, and endless possibilities. Qwen develops open-source and proprietary large language models (LLMs) to advance AI accessibility globally. The name 'Tongyi Qianwen' means 'seeking an answer by asking a thousand questions,' reflecting the mission to make AI universally helpful.
Founding Story
Qwen (Tongyi Qianwen) was launched in April 2023 by Alibaba Cloud as a strategic response to the global AI revolution sparked by ChatGPT. The name 'Tongyi Qianwen' translates to 'seeking an answer by asking a thousand questions,' embodying the vision of making AI accessible to everyone. Under the leadership of 32-year-old researcher Lin Junyang, who previously developed M6, the team aimed to create both proprietary and open-source models to democratize AI access. The initial beta was approved by Chinese regulators and opened for public use in September 2023. The founding vision centered on building a comprehensive AI ecosystem that could serve individual users, developers, and enterprises alike, leveraging Alibaba's massive cloud infrastructure and e-commerce platform integration.
Discussions
No discussions yet
Be the first to start a discussion about Qwen Team (Alibaba)
Leadership
Founders
Jack Ma
Former English teacher from Hangzhou, China. Founded Alibaba Group in 1999 with 17 other co-founders.
Executive Team
Lin Junyang
Tech Lead of Qwen Team / Lead Engineer at Tongyi Lab
32-year-old research fellow who developed Qwen, M6, and Tongyi Qianwen. Lead architect at Alibaba's core AI engineering team. Recently set up a robotics and embodied AI team within Qwen (October 2025).
Eddie Wu (Eddie Yongming Wu)
CEO and Director of Alibaba Group
Appointed CEO in September 2023. Previously Chairman of Taobao and Tmall Group.
Business Model
Revenue Model
Hybrid open-source and proprietary model. Revenue streams include: (1) API usage fees for proprietary flagship models (Qwen-Max, Qwen-Plus) charged per token; (2) Cloud computing services - developers using open-source models require compute power, driving Alibaba Cloud revenue; (3) Enterprise licensing for companies with >100M monthly active users; (4) Integration with Alibaba's ecosystem (e-commerce, travel, payments) to drive transaction fees; (5) Free consumer app (Qwen Chat) with 30 million MAU driving ecosystem engagement. API inference costs have dropped 97% year-over-year as of 2024.
Pricing Tiers
32K context window, flagship model with >1 trillion parameters
256K context window
1M context window, balanced performance model
131K context window
1M context window, high-speed inference
1M context window, optimized for speed
131K context window, via Groq infrastructure
Limited free tier available in Singapore region only. Qwen Chat consumer app is completely free to use.
Pre-purchased savings plans offering discounts on Alibaba Cloud
Asynchronous batch jobs receive 50% discount on Alibaba Cloud
Target Markets
- Technology companies (80% of China's tech companies use Alibaba Cloud)
- Large model/AI companies (50% of China's large model companies)
- Automotive industry
- Gaming industry
- Scientific research institutions
- Financial technology (fintech)
- Chatbot and conversational AI
- Customer service automation
- Code generation and programming assistance
- Mathematical problem solving
- Document analysis and summarization
- Image generation and editing
- Airbnb
- Apple
- 200+ million users via Quark application in China
- DingTalk
History & Milestones
Qwen family surpassed 700 million downloads on Hugging Face
Launched Qwen2.5-VL (Vision-Language) in 3 sizes
Released Qwen2.5-Max, flagship MoE model
Released Qwen3 model family (dense and MoE models)
Released Qwen3-Coder-480B-A35B-Instruct
