Introducing System One Models and Jev

China’s MiniMax LLM costs about 200x less to train than OpenAI’s GPT-4, says company

How a $6M bet on Wiz turned into a massive 200x return for one early backer

How we made a Ruby method 200x faster