This model does not do reasoning.
This model always do reasoning.
Your environment does not support WebGPU acceleration. Try using API models instead.
