OpenAI halves response costs for guest ChatGPT users
OpenAI optimized its AI models, cutting inference costs by over 50% for ChatGPT.
Published 6sem1 source
Lire en français
≈ 19s
The fact
The number of required Nvidia GPUs dropped to a few hundred, down from thousands previously.
These cost savings primarily benefit guest (non-paying) users of the service.
Click the link to read an article on the topic: