OpenAI just handed 100,000 scientists free access to its most powerful AI models. No strings attached. Almost.

The program, called "OpenAI for Science," gives researchers from biology to climate science a bucket of free API credits. Enough to run experiments that would normally cost thousands of dollars.

But here's the question everyone's asking: how can they afford to give away something that valuable?

The short answer is they're not losing money. They're investing it.

OpenAI's real cost per API call is tiny. Once the models are trained, serving a query costs pennies — sometimes fractions of a cent. The massive upfront training cost is already sunk. So giving away credits to 100,000 researchers doesn't blow a hole in their balance sheet. It's a rounding error compared to the billions Microsoft poured in.

But the real payoff is bigger. Every scientist using the platform generates data — questions, prompts, corrections, novel use cases. That data trains the next generation of models. OpenAI gets a free, high-quality feedback loop from the world's smartest people.

And then there's the lock-in. Once researchers build their workflows around OpenAI's APIs, switching costs go way up. Universities and labs will start paying for expanded access when their free credits run out. It's a classic freemium play.

Sponsored Deal

Microsoft's $10 billion investment gave OpenAI the runway . But they're not just burning cash. They're building a moat.

Competitors like Google and Anthropic have similar programs, but none as big. OpenAI's bet is that giving away access now will create a generation of researchers who default to OpenAI. And when their discoveries lead to commercial products, OpenAI gets a cut.

So no, it's not charity. It's a long-term bet on data, loyalty, and market dominance. The question isn't whether they can afford it. It's whether their competitors can afford not to do the same.