Hacker News new | ask | show | jobs
From GPT-4 to Mistral 7B, there is a 300x range in the cost of LLM inference (twitter.com)
2 points by Gcam 859 days ago