Moonshot AI shifts Kimi models from coding to science amid benchmark gains flattening
moonshot ai redirects kimi models from coding to biology, chemistry, and pharma research due to diminishing returns in coding. this pivot aims to leverage smarter ai models for scientific applications.
moonshot ai, a beijing startup, is pivoting its kimi language models from coding to applications in biology, chemistry, and pharma research. according to giacomo ficari, international director, science demands smarter ai models than law, finance, or code. the kimi k3.1 pro model is being integrated into dp technology's research assistant biomaster, where it has shown efficiency in analyzing gene data. in one test, kimi analyzed data from 3,696 mouse pancreatic cells and identified 100 candidate genes in 90 minutes, compared to an estimated four hours of human work. however, the pharma industry remains cautious, with a june citi survey finding that while 72 percent of us research leaders are using ai broadly, only 24 percent expect it to noticeably improve drug development success rates.