I am connecting via the Zapier API and am receiving the subject line message. I’ve checked the size of the input summing the “User Message” & “Assistant Message” and get 1,897 tokens with 9,107 characters.
I have 90K TPM / 3,500 RPM on GPT 3.5 and 10K TPM 3K RPM on GPT4.
Can someone tell me whether it is Zapier or me that is being stupid? Because it doesn’t matter which model I set Zapier tells me I do’t have enough tokens.
Am I missing some part of the calculation?