Reddit: open-source models 'writing on the wall' for Anthropic, OpenAI

A Reddit post argues that the acceleration of open-source models, citing recent releases like Minimax 3 Pro (2.7T parameters) and GLM 5.3, signals a decline in enterprise trust for proprietary AI providers such as Anthropic and OpenAI. The post highlights the 'Kimi moment' as a turning point.