OpenAI launches GPT-6 Sol and Luna, reduces API prices by 50%

ChainCatcher report: On September 22, OpenAI released two new models, GPT-6 Sol and GPT-6 Luna, just minutes after Anthropic launched Claude Opus 5.5. Positioned below GPT-6 Astra, these models offer a more affordable and faster option for everyday tasks. OpenAI also reduced the API pricing for both models by 50% compared to the promotional rate of GPT-5.6: Sol’s input and output prices are $2 and $10 per million tokens, respectively, while Luna’s prices have been lowered to $0.10 and $0.50. In terms of performance, OpenAI cited Zapier’s AutomationBench benchmark, showing that GPT-6 Sol achieved a score of 33.2% at its highest inference setting with a cost of $0.27 per task, compared to Claude Opus 5’s highest setting of 26.9% and a cost over 11 times higher. In the Agents' Last Exam test, GPT-6 Sol scored a maximum of 56.4%. In the OSWorld 2 computer usage test, GPT-6 Sol nearly matched Claude Opus 5’s medium-effort score of 60.3% with a score of 60.5%, at approximately 80% lower cost.