Gpt oss 120b not reasoning all the time

Hi,

I’ve seen that from time to time, the model answers directly without reasoning first. It’s usually after calling a tool but not all the time.

Is it normal? I see it on all of my flows now

1 Like

Could you offer the former model on another endpoint maybe? Like a stable one that wont change. It is critical that I have a stable and reasoning model. My workflows are literally broken with the model that skips reasoning like a third of the time

Can you please answer?