OpenAI Launches GPT-6 New Models at Half Price
OpenAI has announced GPT-6 Sol and GPT-6 Luna. Both models maintain performance equivalent to the previous generation while cutting token-per-cost in half. However, multiple independent analyses found little substantial improvement in intelligence, and the release is evaluated as prioritizing price competitiveness over performance gains.

OpenAI has announced two new models: 'GPT-6 Sol' and 'GPT-6 Luna'. The most significant feature of both models is that they maintain performance equivalent to the previous generation while cutting the processing cost per token in half. This release, centered on price reduction, is positioned as a strategic move prioritizing cost competitiveness over major functional expansion.
In the AI industry, rapid price declines have continued alongside model performance improvements in recent years. As companies have accumulated processing efficiency gains, an environment is emerging where the same quality of service can be provided more affordably. GPT-6 Sol and Luna align with this industry-wide trend of price competition, and are noted as being priced with particular attention to Anthropic's relatively expensive model lineup.
However, performance evaluations require careful consideration. According to multiple independent analyses, substantial improvements in intelligence in GPT-6 Sol and Luna were barely confirmed. In other words, this refresh emphasizes delivering 'the same level of intelligence more affordably' rather than 'becoming smarter,' making it a different type of update from a technological breakthrough.
Worth noting is that Anthropic simultaneously released 'Opus 5.5'. For OpenAI, this same-day release was reportedly unexpected, resulting in overlapping product announcements from both companies. The scenario of competing top-tier AI companies launching new models at the same timing exemplifies how intensified competition has become within the industry.
For users and enterprise developers, this price reduction holds significant meaning. With the same budget enabling more processing capacity, the cost barrier for API-based application development is lowered. Particularly for cost-sensitive user segments like startups and individual developers, favorable conditions for adoption are being established.
However, the assessment that performance improvements are barely evident also has limited appeal to users who base model selection on performance criteria. The presence of GPT-6 Sol and Luna ultimately depends on how widely adoption expands in cost-focused use cases. The performance and price comparison with Anthropic's Opus 5.5 may influence future market choices.
From an industry-wide AI perspective, this release can be seen as symbolizing a shift in competitive focus from 'performance' to 'economic efficiency'. As technological differentiation becomes difficult, how each company strategizes to capture the market will be a critical factor shaping the future AI industry landscape.
This article is an original work independently written and edited by the AI issue editorial team based on factual reporting. © AI issue. Unauthorized reproduction, redistribution, or use for AI training is prohibited.