At a Glance · AI-generated

Alibaba released Qwen3.6-35B-A3B on April 16, an open-source sparse mixture-of-experts vision-language model with 3 billion active parameters designed for agentic coding tasks. The model demonstrated superior performance compared to Google's Gemma 4 on coding benchmarks while remaining lightweight enough to run on consumer hardware. Early testing showed the model producing competitive results against proprietary systems like Claude Opus 4.7 across various capabilities.