>verbose

GLM defaults to max effort btw

https://docs.together.ai/docs/glm-5.2-quickstart#reasoning-e...