OpenAI has officially launched its latest AI model, o1-pro, as part of its developer API. This model is priced at ten times higher than its predecessor, o1, reflecting its enhanced performance capabilities. Initial benchmarking results indicate that o1-pro achieves an output speed of approximately 9 tokens. In a notable achievement, o1-pro set a new record in the Extended NYT Connections benchmark with a score of 81.7, surpassing the previous record held by o1, which scored 69.7. The model is designed for complex problem-solving and offers improved reasoning, making it one of the most advanced offerings in the AI landscape. Additionally, the model is available on OpenRouter, where it is characterized as slow but deliberate in its processing. The launch of o1-pro marks a significant advancement for OpenAI in the competitive field of large language models.