I've been using Kimi k2.6 extensively via kimi-code and I only reach for frontier models when I do a multi-model security review (and Kimi actually does a better job of finding stuff, albeit with more false positives -- I often run Kimi's output through Opus 4.7/8 and Opus will concur that Kimi found genuine issues, while Opus didn't actually find those issues itself, for example).

So whatever, I just don't really feel the need to burn tokens on Fable anyway.