Political-bias benchmark for Grok 4, GPT-4.1, Gemini 2.5 Pro and Claude Opus 4 promptfoo.dev 3 points by dangelosaurus 5 months ago · 0 comments Reader PiP Save No comments yet.