OpenAI's latest flagship model is **o1**, released in December 2024. 

The o1 model represents a significant advancement, featuring enhanced reasoning capabilities and trained using reinforcement learning to "think before it answers." It comes in different variants, including o1 (the full version) and o1-mini (a faster, more cost-effective option).

Before o1, GPT-4o (released in May 2024) was the flagship, offering multimodal capabilities across text, vision, and audio.
