[staticman2]

I've seen Perplexity misrepresent search results and also interpret them differently depending on whether GPT4o or Claude Sonnett 3.5 are being used.