Alibaba’s Qwen Team has unveiled the Qwen3.5 Small Model Series, which includes the remarkably efficient Qwen3.5-9B, capable of outperforming OpenAI’s gpt-oss-120B while being 13.5 times smaller. This release comes amid ongoing political instability in the U.S. AI sector, highlighting China’s continued advancements in AI technology. The models are built on a hybrid Efficient Hybrid Architecture that enables higher throughput and lower latency for tasks, supporting a range of local applications from laptops to smartphones. Available under the Apache 2.0 license, these models can be used and modified commercially, providing developers significant flexibility and accessibility in deploying AI solutions without reliance on cloud infrastructures.
Alibaba’s Qwen3.5-9B outperforms OpenAI’s models on benchmarks
