Look at the logo behind Alan Trefler…
"Token Co$t" with a line through it. For me, that was one of the biggest announcements at
#PegaWorld yesterday.
For the first time, an enterprise software company is offering agentic AI with zero token charges. Clients pay a flat price per completed case, per outcome, rather than per token. Things like a claim processed, a customer request resolved, or an employee onboarded.
Honestly, I did not see this coming. I assumed the change to the token model would come from a startup, not from one of the most established players in the industry.
So the obvious question is, how can
@Pegasystems afford to do that?
@Alan Trefler explained it with a kitchen. A great restaurant does not redesign the menu with every order. The chef perfects the recipe once in the test kitchen, the team follows it, and the result is the same dish at predictable cost, every time.
That is when it clicked for me. Most enterprises are running AI the opposite way. Every customer interaction triggers a new reasoning loop, and the model rethinks the answer to a question it has already answered a thousand times. That is where the token bill explodes.
And here is the part that struck me most. The shift is not only about cost. It is also about who carries the risk. When you pay per token, every inefficiency in the architecture lands on your invoice. When you pay per outcome, the vendor has to make the architecture efficient, or they lose money. The incentive flips.
This may quietly reshape enterprise AI procurement over the next 18 months. Pricing AI by activity will start to look like paying for dial-up by the minute.
What part of your AI stack is still billed by the meter?
#PegaPartner #AgenticAI #AIReadiness #EnterpriseAI #FutureOfWork