Kimi-K3.1-pro vs DeepSeek V4.1 Flash: DeepSeek still ahead on 3D
A user comparing Kimi-K3.1-pro with DeepSeek V4.1 Flash 1017 says the gains are not profound, DeepSeek Flash remains stronger at 3D, and Kimi does not fix the areas where DeepSeek fails.
Initial impression of Kimi-K3.1-pro is really good, but when compared with DeepSeek-v4.1-flash-1017 there is not that profound improvements. 3D capabilities are not that high, i would say deepseek-flash still has the upper hand on it. But Kimi can be guided to get some really good results, but so does deepseek. Other way to put this is, Kimi K3.1-pro is not performing well in areas Deepseek-flash is failing, which i hope it did. Kimi-K3.1-pro is good, not saying it's not, but on some aspects i'm liking deepseek flash result more. It's also 3T vs half a T, so you expect some major improvements. True test needs api access, so far all Kimi test is via Kimi code.
This post quotes another post: x.com/itsmeajaykv/status/2113755557182267586