Chinese Academy of Sciences Launches 'Panshi 100' Model System

'Panshi 100' Model System Officially Unveiled, Ushering in the Era of AI-Powered Research Platforms

On April 28, the Chinese Academy of Sciences (CAS) officially launched the 'Panshi 100' model system. The introduction of this system marks a pivotal shift in China's AI-driven scientific research — moving from fragmented, isolated explorations toward a collaborative and efficient platform-based innovation paradigm, providing robust intelligent support for cutting-edge research across various fields.

Three-Layer Architecture: Full-Chain Coverage from Foundation Models to Specialized Applications

According to official sources, the 'Panshi 100' model system adopts a clearly structured three-layer architecture:

• Base Layer — Panshi Scientific Foundation Model: Serving as the core foundation of the entire system, this base model aggregates cross-disciplinary scientific knowledge and reasoning capabilities, providing general-purpose intelligent support for upper-layer models.
• Middle Layer — Discipline-Specific Models: Dedicated vertical models have been developed for specialized fields including physics, chemistry, biology, materials science, and earth sciences, each equipped with deep domain understanding and professional reasoning capabilities.
• Top Layer — Scenario-Specific Models and Intelligent Tools: Targeting specific research scenario requirements, a series of specialized application models and supporting intelligent tools have been developed to directly serve the daily work of researchers.

Together, these three layers form a broad-coverage, highly collaborative intelligent research platform, achieving full-chain integration from foundational capabilities to practical applications.

From 'Isolated Breakthroughs' to 'Platform Collaboration': A Paradigm Shift in AI-Driven Research

For a long time, AI applications in scientific research have mostly proceeded in an 'isolated exploration' fashion — individual research groups training specialized models for specific problems, with little collaboration or reuse between them. While this approach has yielded notable breakthroughs in localized areas, overall efficiency has been limited, making it difficult to generate systematic synergies.

The launch of the 'Panshi 100' model system aims precisely to break this fragmented landscape. Through a unified foundation model base, discipline-specific models across different fields can share underlying capabilities. Through standardized interfaces and toolchains, upper-layer applications can be rapidly built and iterated. This platform-oriented approach is expected to significantly lower the technical barriers to AI-powered research, enabling more researchers to conveniently leverage artificial intelligence to accelerate their work.

CAS's AI Research Strategy Carries Far-Reaching Significance

As China's premier academic institution in the natural sciences, the Chinese Academy of Sciences' launch of the 'Panshi 100' model system sends multiple positive signals. On one hand, it demonstrates that China's strategic positioning in the AI for Science domain is transitioning from concept to systematic practice. On the other hand, coordinating efforts through a national-level research institution helps consolidate superior resources, avoid redundant development, and accelerate the creation of internationally competitive AI research infrastructure.

Outlook: A Promising Future for Intelligent Scientific Research

Currently, the deep integration of AI and scientific research has become an irreversible global trend. From DeepMind's AlphaFold to Microsoft's scientific computing platforms, international tech giants are ramping up their investments. The launch of the 'Panshi 100' model system provides China with critical platform-level support in this competitive arena.

Looking ahead, as more discipline models are integrated and application scenarios expanded, 'Panshi 100' is poised to become a key piece of infrastructure driving the intelligent transformation of scientific research in China, empowering researchers to achieve innovative breakthroughs across broader frontiers.