OpenAI tests their add-on strategy

Recently I have noticed OpenAI ratcheting up its A/B testing on its site.

This recent test is not particularly novel: having add-ons above your normal pricing plans loses to having them below.

However, I am intrigued to see how OpenAI approaches its pricing strategy in 2026.

SaaStr summarized their State of AI report and noted "Enterprises aren’t just chatting with AI. They’re running complex, multi-step reasoning tasks in production. 320x increase in reasoning token consumption per org YoY means this went from experiment to infrastructure."

It feels like it might be the B2B side, and not the consumer side, which is the major revenue lever that OpenAI leans into.

0 comments