Moonshot AI Pivots Kimi From Coding Benchmarks to Science Research
The Decoder · rss · 2026-11-17
Moonshot AI says it is done chasing coding benchmarks and is pivoting its Kimi language models toward biology, chemistry, and pharma research. The Beijing startup argues science demands smarter AI than law, finance, or code—a shift coming as gains on coding benchmarks flatten across the industry.
More from Companies & People
- Founder: OpenAI faces an existential moment unless it ships a new model this week — bindureddy · 2026-11-17
- Artificial Analysis to host image/video model benchmarking event, having tested 200+ image and 130+ video models — ArtificialAnlys · 2026-11-17
- Polymarket odds hit 42% for a fourth Jensen Huang affair report by year-end — Polymarket · 2026-11-17
- Sam Altman says OpenAI "will go public at some point" but warns of real downsides — ChrisGPT · 2026-11-17
- Claude reportedly trailing Copilot in adoption within Microsoft Office apps — Polymarket · 2026-11-17
- VC: AI spend now big enough that enterprises mix and match multiple model vendors — pwendell · 2026-11-17