MiniMax says M3.1 Flash makes online inference nearly three times faster and cuts usage costs 70%; its coding and agent ...