2 ms·this is a good model. I benchmark reasoned answers to qwen 3.6 27b (no think)+ bash and it held up.by unfirehose 3mo agothis is a good model. I benchmark reasoned answers to qwen 3.6 27b (no think)+ bash and it held up.