update on monkAI
Got advice from an anon. He said to do some nerd optimisations shit.
I made claude do it. Now my qwen 3 coder went from like 16 tokens a second to 170 a second and uses like half the amount of tokens it was using before.
Based and redpilled.