OpenAI GPT-5.5 Models Surpass Junior Lawyers on Legal Research Benchmark

In my view, it is correct to say that recent OpenAI models (such as GPT-5.5 Pro and GPT-5.5 Thinking (Heavy)) outperform many (most?) good junior lawyers on the particular legal research questions that comprise the Legal Research portion of my benchmark. Of course, a lawyer's job consists of much more than just legal research.