An internal tool classifies incoming messages into one of eight categories at high volume. The team is using the largest available model and cost has become the main objection. What is the most promising first change?

• AIncrease the context window so fewer calls are needed.
• BEvaluate a smaller model on the same task, since narrow classification rarely needs the most capable model.
• CRaise the temperature so the model reaches an answer in fewer tokens.
• DFine tune the largest model so it responds more briefly.