AI is changing the world. Don't stay behind. Clear summaries, community insight, delivered without the noise. Subscribe to never miss a beat.

Something is afoot in the land of Qwen

simonwillison.net

March 4, 2026

4 min read

🔥🔥🔥🔥🔥

72/100

Summary

Alibaba's Qwen team has released the Qwen 3.5 family of open weight models. Junyang Lin, the lead researcher, announced his departure from the team via Twitter.

Key Takeaways

Alibaba's Qwen team released the Qwen 3.5 family of open weight models, including sizes ranging from 0.8B to 397B, with notable performance in coding tasks.
Junyang Lin, the lead researcher for Qwen, announced his resignation amid a company reorganization, triggering additional departures from the team.
Alibaba's CEO addressed employees in an emergency meeting, indicating the company's awareness of the significance of the recent resignations.
The Qwen 3.5 models have received positive feedback for their effectiveness, particularly the smaller models, which are capable of multi-modal reasoning.

Read original article

Community Sentiment

Mixed

Positives

Qwen3.5 is an impressive model, demonstrating strong capabilities in coding tasks with minimal guidance, which could enhance developer productivity.
The model's ability to write tests that pass without cheating indicates a high level of reliability and understanding of coding principles.

Concerns

Some users report disappointing experiences with the new Qwen model, suggesting it may not effectively utilize standard coding tools, raising concerns about its practical utility.
There are worries that the models might be overfit to evaluation metrics, which could limit their real-world applicability and effectiveness.