Chinese AI Firms Pivot Toward Scientific Research Automation
Moonshot AI and Alibaba DAMO Academy are prioritizing scientific discovery as a key benchmark for evaluating the capabilities of their latest AI models.
Chinese artificial intelligence companies, including Moonshot AI and Alibaba DAMO Academy, are increasingly centering their development efforts on scientific discovery and research automation. These firms are positioning specialized scientific capabilities as a primary competitive benchmark for their latest AI models.
According to aitimes.com, this shift is exemplified by recent integrations and research initiatives. On November 12, 2026, DP Technology integrated the Kimi K3.1 Pro model into its BioMaster AI agent. The system demonstrated its utility by analyzing gene expression data from 3,696 mouse pancreatic cells in 90 minutes.
Several other milestones in the sector have preceded this focus. In July 2026, Alibaba DAMO Academy announced that its AI agent, Elements Claw, discovered four new superconducting materials. That same month, the Chinese Academy of Sciences released ScienceOne Omni 2.0, a model designed specifically for scientific applications. Additionally, on November 5, 2026, Moonshot AI founder Yang Zhilin and Andrew Yao released the TCSExam research, which evaluates AI models across 100 distinct research tasks.
Quick answers
What is the TCSExam?
Released on November 5, 2026, by Yang Zhilin and Andrew Yao, the TCSExam is a research evaluation designed to assess AI models across 100 research tasks.
What did the BioMaster system achieve?
On November 12, 2026, the BioMaster system, using the Kimi K3.1 Pro model, analyzed gene expression data from 3,696 mouse pancreatic cells in 90 minutes.