Alibaba Qwen Team Advances AI Models with Qwen3.6-Plus and Dense Qwen3.6-27B Releases

Alibaba's Qwen team has accelerated its model release cadence with two new offerings in April 2026. On April 1, they unveiled Qwen3.6-Plus, a closed-source model that significantly improves agentic coding capabilities—allowing the AI to autonomously plan and execute multi-step software development tasks. The model also sharpens multimodal reasoning (e.g., image-to-code, chart analysis) while supporting a 1M-token context window, enabling processing of entire codebases or lengthy documents in one pass.

On April 22, the team released Qwen3.6-27B, a dense open-weight model under Apache 2.0 license. Its headline feature is a Thinking Preservation mechanism that stores and reuses reasoning traces from earlier steps, boosting performance on long-horizon agentic tasks without proportional compute increase. Together, these releases signal Alibaba's strategy to deliver state-of-the-art AI that runs efficiently on local devices—from laptops to edge servers—democratizing advanced agentic workflows beyond cloud-only setups.