📰0062.8% on Aider Polyglot from a MacBook Pro. Then the other model we tried scored 4%. Here's what actually happened, with a working cost loop attached.DEV Community·Christopher Maher·about 1 month ago#PWkpNTLn#infercostv030#llmkubev072#kubernetes#ai#aider#infercost+5 more🧰Tag tools✨Add tagQwen3.6-35B-A3B Q8 on a MacBook Pro M5 Max scored 62.8% on Aider Polyglot, beating Claude Sonnet 4 with 32k thinking. Then Devstral 2 scored 4% on the same harness but 81.7% on HumanEval+.… Read more15s0Read later0Read More
