In terms of openness, Kimi is much freer, and you can create your own prompt words (common phrases) and DIY gameplay. For other large models, the official basically gives you a lot of intelligent agents, and you find and use them yourself, instead of letting you build your own intelligent agents.
In terms of writing style, Kimi is slightly inferior to Tongyi Qianwen (Qwen), but not inferior to Wenxin Yiyan (ERNIE Bot).
Online search and long text processing are Kimi's advantages, which are difficult for other large models to achieve. Recently, Tongyi Qianwen (Qwen) updated the online search and traceability functions, and is learning Kimi.
In terms of logic and reasoning, everyone is about the same, and has not yet caught up with GPT-4.
Zhipu AI focuses on the B-side and does not seem to be very interested in the C-side.