AI tool pricing rests on two different models. The first is a per-user, per-month licence, which is predictable and suits employees' regular work. The second is pay-per-use, where the volume of processed text is billed converted into tokens – the units into which both input and output are broken down – or credits for an image, a minute of video, or a transcription. For automations and integrations, the second model tends to be more advantageous, but the cost grows with volume and without a limit it can spring an unpleasant surprise. The budget should therefore include account caps, alerts when they are exceeded, and an overview of how much each team consumes. When comparing offers, it is worth calculating the cost per task completed, not the licence price, because that is what can actually be compared with the work of a person.
See also: Trial versions and pilot licences, How much does an AI solution cost, Return on AI investment.
