how does deepseek-r1 compare to other reasoning models

Back to top