[findjashua]

did a comparison on LM Studio - the answers are eerily similar, but Mixtral is way way faster. Codellama-70B is slow to the point of being unusable.

(M1 Max, 64 GB RAM)