The artificial intelligence market is entering a phase of intense price war. On July 30, OpenAI announced a radical price cut for two models in the GPT-5.6 line: the Luna tier was slashed by 80%, and Terra by 20%. This move, made just three weeks after launch, looks like a direct response to growing pressure from corporate clients who are increasingly scrutinizing their AI solution budgets.

Against the backdrop of a global economic slowdown and tightening corporate budgets, companies are no longer willing to pay for "raw" technology. Chinese competitors, such as Kimi K3 from Moonshot AI and GLM-5.2 from Z.ai, offer comparable solutions at significantly lower prices, forcing the American giant to reconsider its pricing policy.

Details of the new price list

The new rates are as follows: for the Luna model, token input costs dropped from $1 to $0.20 per 1 million tokens, and output from $6 to $1.20. For Terra, input prices fell from $2.50 to $2, and output from $15 to $12 per 1 million tokens. The flagship Sol model retained its previous rates, underscoring its premium status.

This is no longer just an adjustment, but full-scale dumping. The price cuts follow years of growth in corporate AI spending, when employees tested technologies without regard for budgets. Now CFOs demand measurable returns on investment, and OpenAI has to adapt.

True motives: efficiency or panic?

The company insists that the price reduction is the result of increased efficiency, not a forced measure. "We continue to expand capacity while achieving efficiency. Each new generation of AI should perform more tasks at lower cost," OpenAI states.

However, the timing is no coincidence. Chinese models have significantly strengthened their positions this year, offering solutions cheaper than competitors. Even Anthropic's mid-tier model, Claude Sonnet 4.6, costs more than OpenAI's updated Terra. This suggests that American companies are beginning to lose the price race to Eastern competitors.

Analysts see a double game in this move. On one hand, dumping can attract new clients and expand the user base. On the other, it directly hits margins, which is critical ahead of the initial public offering (IPO) that both companies are preparing for.

Impact on IPO: playing ahead of the curve

The price cut will have a twofold impact on IPO preparations. Broad audience reach makes the brand more attractive to future shareholders, and revenue dynamics and user base growth now matter more than net profit. Retaining clients protects the company from audience outflow to Eastern rivals.

However, investors will eventually demand a clear plan for achieving stable profitability. The margins of basic services are falling along with the cheaper tokens, complicating the creation of an attractive securities prospectus. If the volume of requests to Terra and Luna does not grow sufficiently, financial metrics will suffer, and this will be reflected in reports as a decline in overall revenue.

It is still hard to predict whether dumping will save the situation or merely delay the drop in revenue. OpenAI's next financial report should reveal the first results of the new strategy.

My view: This is a classic example of a "growth for growth's sake" strategy at the expense of profitability. In the short term, such a move may protect market share before the IPO, but in the long term, it sets a dangerous precedent for a price race in which only those with lower computing costs win. Chinese companies, with their government support and cheap energy, look like the favorites in this race.