Something is afoot in the land of Qwen

2026-03-04 · Source: Simon Willison's Weblog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, short

Summary

Alibaba's Qwen team is experiencing significant upheaval, with lead researcher Junyang Lin and several other core members, including Binyuan Hui and Bowen Yu, announcing their resignations on March 4th, 2026. This exodus, potentially triggered by a re-organization placing new leadership from Google's Gemini team in charge of Qwen, prompted an emergency "All Hands meeting" with Alibaba Group CEO Wu Yongming. The departures are particularly impactful as they follow the recent release of the "exceptionally good" Qwen 3.5 family of open-weight models, ranging from 397B to 0.8B parameters, which have shown remarkable quality, especially the smaller versions for coding and multi-modal tasks. The situation leaves the future of the Qwen project uncertain, despite its proven track record in developing high-quality, efficient AI models.

Key takeaway

Alibaba's Qwen team, responsible for the highly effective Qwen 3.5 open-weight models, is experiencing significant leadership departures, including lead researcher Junyang Lin and core contributors to Qwen-Coder and Qwen-Instruct. This exodus, potentially triggered by an internal re-org, casts uncertainty on the future of Qwen's continued innovation in delivering high-quality, smaller-scale models like the 2B multi-modal variant. AI/ML professionals relying on Qwen's proven track record for efficient, performant models should monitor this situation for potential impacts on future releases or new ventures from the departing talent.

Topics

• Qwen 3.5
• Alibaba
• AI Research Leadership
• Open-weight Models
• Large Language Models

Best for: AI Architect, Machine Learning Engineer, NLP Engineer, AI Researcher, AI Engineer, Director of AI/ML

Related on AIssential

• Alibaba just open-sourced a model that rivals GPT-5.2 — Alibaba's Qwen 3.5, with 17 billion active parameters, rivals top closed-source models while being open-weight.
• Alibaba launches Qwen 3.5 AI model amid intensifying China chatbot race — Alibaba's Qwen 3.5 model offers significant cost reduction and workload capacity for agentic AI tasks.
• 🗞️ Alibaba released their open-source Qwen3.5-9B model, and it holds its own against OpenAI's 120B param gpt-oss system. — Smaller, efficient AI models are achieving performance comparable to much larger systems through architectural innovations and multimodal training.
• Qwen 3.5 Small Models: Alibaba Delivers Compact Multimodal Powerhouses — Alibaba's Qwen 3.5 Small models offer powerful multimodal AI for edge devices.
• Alibaba's Qwen 3.5 397B-A17 beats its larger trillion-parameter model — at a fraction of the cost — Sparse Mixture-of-Experts models can surpass larger predecessors in performance and efficiency.
• Alibaba's new open source Qwen3.5-Medium models offer Sonnet 4.5 performance on local computers — Alibaba's Qwen3.5 models offer high-performance, open-source LLMs with efficient local deployment and competitive API pricing.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Simon Willison's Weblog.