OpenAI launches new super-efficient, low-latency ‘GPT-4o mini’ [cointelegraph]

The new model purportedly offers comparable performance at a fraction of the cost.