I made ChatGPT, Claude, and Gemini summarize a report I'd actually read, and caught all three making things up
…For Claude, I would use Claude Opus 5 with at least high effort. Anthropic also provides Extra High and Max settings, with Max giving the model the greatest reasoning depth. In Gemini…
