4 ms·OpenAI real-time api, vision fine-tuning, prompt caching, model distillation2 points by yayitswei 2y ago