Why Is OpenAI Suddenly This Cheap?

Ok so here is the thing nobody says out loud. A company that is reportedly losing tens of billions of dollars a year just made its cheapest model even cheaper. Twice, in the space of a few months. That is not the behavior of a business protecting its margins. That is the behavior of a business protecting something else entirely.

The race everyone thought they were in

For a while the story was simple: whoever builds the smartest model wins. Billions poured into training runs, endless talk of the next generation and eventually something like AGI. Meanwhile a completely different race was happening in the background, and it was not about intelligence at all. It was about who actually gets used, every day, by real developers building real products.

And on that second scoreboard, the cheap, fast, good enough Chinese models had quietly been winning for a long time. Not because they were smarter. Because a developer building a chatbot or an app does not need the smartest model in the world. They need one that is cheap enough to run at scale without thinking twice.

And here is the part that should actually worry the labs betting everything on being first to AGI: what if intelligence was never the bottleneck? What if the whole time, the thing that decided who wins was just... price?

That is basically what appears to be happening now. A frontier lab that has spent years talking about the race to build the smartest possible system just cut its cheap model's price in half, then in half again, until it landed close to what its Chinese competitor charges. That only makes sense if the real fight was never about intelligence. It was about habits. Once a developer wires their product around a cheap provider, ripping that out later is expensive and annoying, even if a rival gets slightly cheaper down the line. Getting back in the door before that habit sets in is worth losing money over.

So is it working?

Maybe, maybe not. It puts the lab back in the conversation, which is not nothing. But it does not erase the deeper mismatch: an enormous amount of money is still going into chasing the frontier, while the actual volume of everyday AI use, the boring stuff that pays the bills, keeps drifting toward whoever is cheapest and good enough. A price cut buys attention. It does not automatically buy the habit back.

Being cheap again does not undo months of someone else being the default. It just means you are allowed back in the room.