Live data from Hacker News

Benchmarking LLM APIs – OpenAI, Cohere or Anthropic

Post reply on HN

Print view View original

1–3 of 3 posts

Benchmarking LLM APIs – OpenAI, Cohere or Anthropic

#1

Post by nileshbansal » Sun, May 12, 2024, 9:07 PM UTC

Benchmarking LLM APIs – OpenAI, Cohere or Anthropic
workorb.ai

Re: Benchmarking LLM APIs – OpenAI, Cohere or Anthropic

#2

Post by rohitpaulk » Sun, May 12, 2024, 9:52 PM UTC

Looks like GPT-4 is 3x slower on Azure compared to OpenAI?

Re: Benchmarking LLM APIs – OpenAI, Cohere or Anthropic

#3

Post by workorb » Sun, May 12, 2024, 11:42 PM UTC

rohitpaulk wrote: post #2

Looks like GPT-4 is 3x slower on Azure compared to OpenAI?

Yes, on Azure hosted version GPT-4 is slower but GPT-3.5 is faster (compared to OpenAI's own hosting).

Post reply on HN