What happened

Alibaba has introduced Qwen3.8, a new update in its Qwen large-model family, with a clear emphasis on coding and agentic workflows. According to InfoQ AI, the release is tied to a 2.4T-scale foundation and includes a notable demonstration: the model reportedly worked autonomously for 16 days to build Hermes Agent.

Why it matters

An AI agent is a system that can break down a goal, use tools, make decisions and iterate toward a result. In software development, that means going beyond code completion: the model may plan features, write modules, debug errors and refine an application over multiple steps. The Hermes Agent example suggests Alibaba is positioning Qwen3.8 less as a chatbot and more as a productivity layer for software teams.

Industry note

The next race in large models is shifting from benchmark scores to reliable task completion; agentic coding may become one of the first areas where that shift is visible to everyday developers.