It's a decent general model too - I have it plugged up in llm and raycast since August at great speeds. I wish Cerebras would do MiniMax M2 which should be an upgrade and replacement if it was just faster. It would never be as fast as gpt-oss-120 though