lesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costslesson 2 of 4 · read the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costsread the token bill — what your llm feature actually costs
Four prompt shapes, all running on Claude Sonnet 5 ($2/1M input, $10/1M output). Which costs the MOST per call?
Be careful — output is 5× the price of input, so token count alone doesn't tell you the answer.