Sorry if it’s not the best place to ask this, of the models in the image, which is the best for (problem solving)/Coding and the best one for studying (ask LLM concepts) ? My PC build is RX 9060 XT 16GB + I3 12100F + 16 GB DDR4 + llama.cpp with Vulkan backend + Linux Mint.
I gave some math problems to Qwen 3.5 27B and Qwen 3.6 27B and they got all of them right, pretty smart models I would say, but very slow and electricity consuming, they took like 5 mins with my GPU at 120 W to solve a problem. The MoE models ans…