This is actually disgusting. Glm 5.3 flash is 50% off 0.25 on openrouter. It is far worse than Luna at 0.50. At this point, might as well support openrouter models to get access to all openai gpt-6.1. It is not like you don't take a cut already.