Aug
18
Link: A new Artificial Analysis benchmark, focusing on OpenAI's gpt-oss-120b, shows how open-weight LLMs exhibit inconsistent performance across hosting providers (Simon Willison/Simon Willison's Weblog)
1 min read