GLM-4.7

K2.5 and GLM-4.7/-5 were good in my experience, another vote for those.

I find it really surprising that you’re fine with low end models for coding - I went through a lot of open-weights models, local and "local", and I consistently found the results underwhelming. The glm-4.7 was the smallest model I found to be somewhat reliable, but that’s a sizable 350b and stretches the definition of local-as-in-at-home.

Tried it for a bit yesterday on a macOS VM. I told it my local mqtt broker hostname and it figured out I have some relays using tasmota, then told it should remember how to toggle those lights and it did. I used Z.ai GLM 4.7 through OpenRouter as its brain. It’s definitely worth checking it out, but keeping in mind the amount of things it can run by having a whole computer to itself.

This would be wonderful if it is accurate - instead of guesstimating, let people report their actual findings. I can confirm GLM 4.7 is possible on M1 Max and it can do nice comprehensive answers (albeit at 12 min an answer) locally. You can also easily do Mistral7B and OSS 20B and others. Structure it as a way to report accruals, similarly to Levels.xyz for salaries, instead of guestimating.