The artificial intelligence market has entered a phase of fierce price war. On July 30, OpenAI announced an unprecedented price reduction for two flagship models, GPT-5.6: the Luna tier dropped by 80%, and Terra by 20%. This is not just a marketing gesture, but a forced measure dictated by the new reality of corporate budgets.

Just three weeks after the launch of GPT-5.6, the company is revising its pricing policy. The reason is obvious: businesses that previously spent generously on AI experiments now demand measurable returns on every dollar invested. Corporate CFOs are blocking budgets without clear ROI, and OpenAI is forced to adapt to this new discipline.

Numbers that speak louder than words

The scale of the changes is impressive. For the Luna model, input costs fell from $1 to $0.20 per million tokens, and output from $6 to $1.20. The Terra model is now priced at $2 for input and $12 for output (previously $2.50 and $15, respectively). The flagship Sol has retained its previous rates, highlighting its premium status. But the main signal to the market is not the numbers, but the context: Chinese competitors such as Kimi K3 from Moonshot AI and GLM-5.2 from Z.ai offer comparable quality at significantly lower prices, and even the mid-tier Anthropic Claude Sonnet 4.6 now looks more expensive than the updated Terra.

Playing ahead of the curve before the public offering

OpenAI insists that the price reduction is the result of increased efficiency, not panic. "Each new generation of AI must perform more tasks at a lower cost," the company states. However, I see a deeper strategic calculation here. Ahead of an IPO, it is critical for companies to show not short-term profit, but sustainable growth in the customer base and revenue. Dumping is a tool for capturing market share and protecting against the outflow of audiences to Eastern competitors.

But there is a flip side to this coin. If the volume of requests to Terra and Luna does not grow multiple times, financial performance will suffer, which will negatively impact reporting to regulators and potential shareholders. The margins of basic services inevitably decline, complicating the creation of an attractive securities prospectus.

My verdict: this is a risky but the only correct bet. OpenAI is sacrificing short-term margins for long-term dominance. The next financial report will show whether this calculation paid off. The AI market is entering an era where the winner is not the most technologically advanced, but the most accessible product.