OpenAI has released two new artificial intelligence models called GPT-6 Sol and Luna. The company says both models are built using the same technology as its earlier Astra system, but they are designed for different needs. Sol and Luna aim to bring advanced AI capabilities to everyday work while reducing costs and cutting down on mistakes that frustrate users.
According to OpenAI, the new models offer a better balance between intelligence and price. This means businesses and developers can run sophisticated AI applications without spending as much money as they would on the most expensive frontier models. The company also claims these models are less prone to errors, which could make them more reliable for customer service, writing, coding, and complex data analysis.
The launch comes as competition in the AI industry continues to heat up. Rivals like Anthropic, Google, and Meta have all released new models in recent months, each trying to offer faster, cheaper, or safer tools. By releasing two models at once, OpenAI appears to be targeting both casual users who want affordable help and professionals who need high performance for demanding tasks.
Analysts say the move could push more companies to replace older AI systems with these newer options. If the cost savings are real, smaller startups might finally be able to afford cutting-edge AI for their own products and services. However, experts will be watching closely to see if the models truly deliver improved accuracy in real-world tests outside of controlled demonstrations.
OpenAI has not shared every technical detail yet, but broader access is expected to roll out to developers and subscribers in the coming days. The release signals that the company is increasingly focused on making powerful AI practical and affordable for mainstream use rather than simply breaking benchmark records. If Sol and Luna live up to their promises, they could become the default choice for companies that need capable AI without enterprise-level budgets.