Kimi K2.7-Code: open-source coding model with better token efficiency

huggingface.co

328 points by nekofneko 8 hours ago


pizlonator - 3 minutes ago

I just had Kimi K2.7-code rebase my Fil-C OpenSSH patch from 3.3.1 to 3.5.7 with quite bare bones instructions and it seems to have worked.

177KB patch, so it's not a small change. The patch did not apply cleanly initially; the agent had to do nontrivial work.

I just showed it the patch against 3.3.1, what command to use to build, and the path to 3.5.7 along with a link to the documentation of the change (https://fil-c.org/constant_time_crypto).

Note, I use my own coding agent (T800, which isn't public, and was previously well tested and tuned for K2.5).

I think this cost me between $5 and $10 in API usage.

giancarlostoro - 5 hours ago

Reading their modified license terms, it cracks me up, because they've basically remade the MIT to be the MIT + the one clause that the BSD used to have, which didn't care about MAU or revenue, if you used it in a product, they asked you to 'advertise' them basically. Honestly, its a reasonable request.

jdw64 - 5 hours ago

Personally, when I use open code or routers, I feel that beyond a certain level, the models don't make a huge difference to me. Except for expensive and mediocre models like Gemini. In that sense, Chinese models are pretty good. I usually write code in function or method units and then design and assemble them together.

GPT series models are more thorough and better, but I'm not sure if the difference is enormous. It seems to depend on the workflow, but in my opinion, if you are thorough enough, I wonder if there really is a big difference