Fine-tuned version of GPT-4o from August 2024, OpenAI's flagship multimodal model with vision and audio capabilities.