Count the tokens used by a multi-turn conversation and see what share of the model context window they occupy, so you know when to trim or compress history.
接近上限時需截斷或總結歷史。