Tencent Hunyuan Releases Hyra 2.0 for Automated Model Development
Tencent Hunyuan has released Hyra 2.0, an AI scientific research system that shifts the focus from solving isolated problems to creating reusable tools and workflows through a multi-layered agent architecture. This release marks a strategic shift in AI for Science (AI4Science) toward compound intelligence, allowing the benefits of tool upgrades to be repeatedly leveraged across downstream tasks and future research. Hyra 2.0 structures research into Overall, Mission, and Worker layers, complemented by separate storage mechanisms for project states, layered memories, and a shared knowledge base called Books.
## BACKGROUND
AI for Science integrates machine learning with scientific workflows to accelerate discoveries in fields like mathematics, physics, and biology. Multi-agent systems use multiple cooperating AI agents with specialized roles to handle complex, long-running tasks autonomously.