Comment by waterTanuki

I regularly use local LLMs at work (full stack dev) due to restrictions and occasionally I get some results comparable to gpt-5 or opus 4