OpenAI's Newest AI Is Twice as Cheap and Makes Half the Mistakes
Your AI assistant just got cheaper and less likely to be wrong.
OpenAI is expanding its GPT-6 generation with updated versions of its smaller Sol and Luna models. According to the company, the 6 series models will be available at half the cost of the 5.6 series of Sol and Luna — a price drop OpenAI attributes to improvements in caching and inference. Sol is designed for complex tasks like coding, while OpenAI says Luna is more appropriate for clerical work — "high-volume tasks with a clear goal, like summarizing documents, extracting information, or answering quick questions." OpenAI also says its newest models are better at producing factually accurate answers, and notes a lower error rate on coding: "On our internal factuality evaluation... GPT-6 Sol makes about half as many mistakes as its predecessor, reaching Astra-level reliability at much lower cost." As is typical with its releases, OpenAI claims Sol and Luna handle tasks substantially better than Anthropic's top models like Fable and Opus — notably, Anthropic released a new version of Opus 5.5 just 90 minutes before OpenAI's release. The new versions of Sol and Luna are now available in ChatGPT Work and Codex for most paid accounts, and in the ChatGPT API, while Luna will be available in the desktop app and for Free and Go users. OpenAI expects to roll the models out to ChatGPT gradually throughout the day.
- OpenAI's new Sol and Luna models cost half as much to use as the last version — a price cut that could trickle down to apps you already pay for.
- OpenAI says Sol makes about 50% fewer factual mistakes, though that figure comes from its own internal testing, not outside reviewers.
- The cheaper model, Luna, is heading to ChatGPT's free tier, so basic tasks like summarizing documents won't cost you anything.
Why It Matters
Cheaper, more accurate AI should mean better apps for less money — and free users get more for nothing.