AI NEWSLatest AI NewsArticleOpenAI Slashes Prices by 80% for GPT-5.6 Luna, Cost-Effectiveness Surpasses DeepSeek V4 ProPublished in Latest AI NewsTime :Jul 31, 2026Read :4minuteOpenAI officially announced the new pricing plan for the GPT-5.6 series. This price adjustment only applies to the mid-range and lightweight models Terra and Luna, while the flagship GPT-5.6 Sol maintains its original pricing. Luna, which is designed for batch lightweight inference scenarios, has seen the largest price reduction in its history, with both input and output costs dropping by 80%.Before the price adjustment, Luna cost $1 per million input tokens and $6 for output. After the adjustment, input costs have dropped to just $0.2 (about 1.4 RMB), and output costs are now $1.2 (about 8.1 RMB). The balanced version of the same series, Terra, has also been reduced by 20%, with input at $2 and output at $12. The Sol model, which focuses on complex code and multimodal deep thinking, remains at its original price, forming a clear tiered strategy: “low-cost Luna captures the market, high-end Sol ensures profit.“Comprehensive cost-performance superiority, compensating for cost shortcomingsOpenAI simultaneously released an evaluation chart from the third-party organization Artificial Analysis, comparing mainstream lightweight commercial models based on two dimensions: cost per task and comprehensive model scores. Data shows that under the new pricing, Luna’s overall cost-performance ratio fully surpasses DeepSeek V4 Pro when completing tasks of the same standard. It scores higher and costs less in scenarios such as general text creation, multi-turn dialogue, and tool calls.Previously, the biggest shortcoming of overseas models compared to domestic high-cost-performance products was cost. This price cut directly addresses that issue. Industry analysts believe the fundamental driving force behind the significant price drop comes from the continuous low-price pressure from domestic large models—DeepSeek V4 Pro achieves extremely low repeated input costs through caching mechanisms, maintaining low prices during off-peak hours. Combined with multiple domestic companies continuously launching affordable API services, this has been continuously diverting overseas small developers. When the low-price advantage of domestic models is no longer a moat, the next battle in the large model industry may have just begun.