About this app

Price a single model call and scale it to a daily and monthly bill, with the output share broken out. Output tokens usually price at three to five times input, so a verbose answer costs far more than a long question. Trimming a system prompt saves less than asking for a shorter reply.